Targeted Advanced Research for Global Efficiency of Transportation Shipping

Creation Date: 24-12-2013 Revision Date: 15-01-2014

Grant agreement no: 266008

Work package: 8 Task: 8.2 Responsible: AMS Deliverable no. 8.3

Title: Guidelines for energy efficient ships

Security status: PU

TARGETS

1/94

ABSTRACT This document covers following contents: Based on task 8.2, this document focuses on producing guidelines for using the TARGETS technologies when defining and designing energy efficient ships for different application and services. Having the comparison results concerning design (Task 5.4.1), retrofitting (Task 6.2.1) and operation (Task 6.2.2) from previous Work packages available, the participants in this task will summarise the findings of the project in a concise guide. This document will build on the description of all alternative configurations for the three phases of a ship’s life cycle produced so far (i.e. construction, retrofitting and operation) and will highlight the most critical ways and configurations that will lead to improved energy efficiency of shipping. The operators participating here will elaborate on the definition of a set of criteria against which each configuration will be compared. In this way, this guide will not only provide a qualitative comparison among various alternatives but also a quantitative assessment of various criteria

TARGETS

2/94

Document Meta Data Document Name: Guidelines for energy efficient ships Document Author/s: Nikolaos Tsakalakis et.al. Document Editor/s: Philip Tsichlis Date of delivery: 04-04-2014 Nature of Deliverable : Report Prototype Demonstrator Other Dissemination Level : PU Security Status: PU Document Status Draft

Final Issued to EC

Partners involved No. Organisation

short name Organisation full name Name email

1 HSVA Hamburgische Schiffbau-Versuchsanstalt GmbH

2 UNEW University of Newcastle Upon Tyne 3 USG University of Strathclyde 4 Maersk A. P. Møller– Maersk A/S 5 AMS Alpha Marine Services Ltd. 6 CMT Center of Maritime Technologiese.V. 7 ITU IstanbulTeknikUniversitesi 8 TUHH TechnischeUniversitätHamburgHarburg 9 Marterra SYNTHESIS MARTERRA AE. 10 S@S Safety at Sea Ltd. 11 SSA Shipbuilders and Shiprepairers

Association

Document history Version Date of

delivery Changes Author(s)

Editor(s) Reviewed by

001 15-1-2014 Initial Draft Nikolaos Tsakalakis Philip Tsichlis

Jochen Marzi

002 04-04-2014

Final Draft Nikolaos Tsakalakis Philip Tsichlis

Jochen Marzi

003 30-04-2014

Final Nikolaos Tsakalakis Philip Tsichlis

004

-- COPYRIGHT 2010-2014 The TARGETS Consortium. This document may not be copied, reproduced, or modified in whole or in part for any purpose without written permission from the TARGETS Consortium. In addition, to such written permission to copy, acknowledgement of the authors of the document and all applicable portions of the copyright notice must be clearly referenced. All rights reserved.

TARGETS

3/94

CONTENTS ............................................................................................................................... PAGE 1 EXECUTIVE SUMMARY ..................................................................................................... 4 2 INTRODUCTION ................................................................................................................. 5 2.1 ENERGY SAVING POTENTIALS (ESP) – GENERAL ................................................ 5 2.2 ESP CATEGORISATION ............................................................................................ 6 2.3 ESP APPLICATION COST ......................................................................................... 6 2.4 ESPS FOR CONSTRUCTION .................................................................................... 7

2.4.1 Short description of selected construction ESPs ................................................... 8 2.5 ESPS FOR RETROFITTING ......................................................................................12

2.5.1 Short description of selected retrofitting ESPs ......................................................14 2.6 ESPS FOR OPERATION ...........................................................................................19

2.6.1 Short description of selected operational ESPs ....................................................21 3 TECHNOLOGIES ADDRESSED AND DEVELOPED IN THE PROJECT...........................24 3.1 IMPROVED CFD METHODS FOR RESISTANCE PREDICTION ..............................24

3.1.1 Overview ..............................................................................................................24 3.1.2 Results .................................................................................................................25

3.2 IMPROVED HULL FORMS ........................................................................................27 3.2.1 Bulbous Bow Optimisation for Slow Steaming ......................................................27 3.2.2 Multihulls ..............................................................................................................29 3.2.3 Aft body optimisation ............................................................................................31 3.2.4 Optimised appendages ........................................................................................34

3.3 IMPROVEMENT OF PROPULSION ..........................................................................35 3.3.1 Design of optimum screw propeller ......................................................................35 3.3.2 Fixed swirl-propeller devices ................................................................................36 3.3.3 Contra-rotating propellers .....................................................................................38 3.3.4 The boundary-layer alignment device (BLAD) ......................................................38

3.4 SAVINGS THROUGH OPERATION – EFFECT OF TRIM AND APPENDAGES .......40 3.5 ASSESSMENT OF ENERGY SAVINGS THROUGH REDUCED RESISTANCE .......42 3.6 VISCOUS RESISTANCE ...........................................................................................45

3.6.1 Surface roughness due to biofouling ....................................................................46 3.6.2 Patterned surfaces ...............................................................................................49 3.6.3 Air lubrication .......................................................................................................52

3.7 ADDED RESISTANCE ...............................................................................................57 3.7.1 Added resistance due to waves ............................................................................57 3.7.2 Added resistance of superstructures ....................................................................59

3.8 RENEWABLE SOURCES FOR PROPULSION .........................................................62 3.8.1 Sails .....................................................................................................................62 3.8.2 Flettner rotors .......................................................................................................69

3.9 AUXILIARY POWER GENERATION .........................................................................71 3.9.1 Photovoltaic installations ......................................................................................71 3.9.2 Fuel cells ..............................................................................................................73

3.10 ENERGY STORAGE FOR MARINE APPLICATIONS................................................79 3.10.1 Necessity of energy storage .................................................................................79 3.10.2 Types of energy storage devices ..........................................................................79

3.11 INTEGRATION OF ENERGY MODULES ..................................................................82 3.11.1 Dynamic Energy Modelling ...................................................................................82 3.11.2 Energy systems modelling ...................................................................................83

4 CONCLUSIONS .................................................................................................................88 5 BIBLIOGRAPHY AND REFERENCES ...............................................................................89 6 INDEXES ............................................................................................................................90 6.1 INDEX OF TABLES ...................................................................................................90 6.2 INDEX OF FIGURES .................................................................................................91

TARGETS

4/94

1 Executive summary Although international maritime transport is the most energy-efficient model of mass transport and only a modest contributor to global CO2 emissions, further improvements are being actively sought as sea transport is predicted to continue growing significantly in line with the world trade and the attendant growth of the world merchant fleet. As far as the optimisation of ships’ fuel consumption and the improvement of energy efficiency are concerned, promising solutions are emerging. Saving energy and thus reducing emissions is one of the few areas in design where the objectives are not contradicting. Namely, reducing fuel consumption is both beneficial for the environment and the operational costs since saving fuel does save money. The complication lies with the choice of the most appropriate means. In this document, the various means available of improving energy efficiency are presented and analysed. Most importantly, a categorisation is presented according to the stage of a vessel’s life that they can be applied at. Finally, their effectiveness and cost will be presented according to the latest developments.

TARGETS

5/94

2 Introduction

2.1 Energy saving potentials (ESP) – General Energy savings can be achieved in a multitude of ways. From energy loss minimisation to waste recovery there are many ways in which one to increase the efficiency of a vessel without compromising its operation. However not all means are suitable for every vessel and thus a systematic study is needed in a case by case basis in order to find the best solutions. Some energy saving potentials will have better results in container vessels while others are more suited to fuller vessels like bulk carriers and tankers. To this end, this study will aim at categorising the various identified ESPs with respect to their suitability for various applications and effect. An example is given in the following figure regarding the use of energy in a small cargo vessel. Although this is a random selection for one ship type, the situation is representative for all ship types. Proportions will change depending on the type of vessel and operational profile. The common denominator is that ultimately all energy from the bunkers will be transformed in heat some of it will first produce useful work in the form of some kind of motion. Thus it is of vital importance not only to identify which losses can be minimised but also which parts of the heat generated can be re-used for various vessel functions in order to achieve better efficiency. For instance, exhaust gases can be used for steam generation.

Figure 1: Use of propulsion energy (IMO, 2009)

TARGETS

6/94

2.2 ESP categorisation To aid application, a categorisation of the various ESPs in hand shall be attempted in this chapter. As a starting point, there are four principal strategies to reduce emissions from ships based on the consumer addressed:

1. Reducing ship resistance 2. Improving propulsion efficiency 3. Improving power generation (machinery) 4. Improving operation

Although propulsion and resistance should be regarded as a single system, their decomposition aids research especially since savings are often cumulative. One drawback though is that the various saving devices or practices may affect the same energy loss. Being mutually exclusive it means that the total gain could be lower than the sum of the individual elements. With this in mind, emission savings can be achieved by addressing these four categories exclusively. It is possible to further separate various ESPs of the three former categories into sub-categories according to whether it is feasible to be retrofitted or can only be applied in construction stage. Not that it is impossible but in most cases the cost makes it prohibitive. For example while it is possible to fit a vessel with a hydrodynamic aid sometime in its lifetime, it is impracticable to change the aft body to an asymmetric design.

2.3 ESP application cost ESP application varies from extremely expensive to virtually cost-free. Most operational measures cost nothing to apply, thus making them very attractive to apply even if they improve efficiency very slightly. On the other hand ESPs that require some sort of modification, either in hull (in the form of an appendage or shaping) or in machinery, have a vast variation in cost. This renders the most expensive ones virtually only applicable in the construction stage where they will only replace another technology and the difference in cost, if any, will only be paid. Those with moderate or little cost usually have a very small investment return period which varies from a maximum of three years to as little as a few months, making them very attractive for retrofitting.

TARGETS

7/94

2.4 ESPs for construction In this chapter, the solutions suitable mostly for application in the construction stage are presented. These are summarised in the following table. As seen they have been separated according to the categorisation presented in the beginning of the document, i.e. Resistance, Propulsion (Hydrodynamic Aids) and Machinery ESPs, while an additional category has been inserted for alternative fuel consideration which could be considered as machinery but has been separated for clarity purposes. Many of these can be applied together but, as underlined earlier, their effects cannot be expected to be added up due to being mutually exclusive. The following table presents some representative examples of ESPs that are mostly appropriate for fitting in the construction stage. The list is definitely neither exhaustive nor exclusive. Some examples of the ESPs in the list are analysed in the following sub-sections.

No. ESP HYDRODYNAMIC AIDS

1 Contra rotating propellers 2 Asymmetric hull aft body

RESISTANCE MINIMISATION 3 “Axe-Bow” shape for tankers and bulkers 4 Minimisation of aerodynamic drag 5 Bare hull optimisation

MACHINERY 6 De-rated M/E 7 Electronically controlled M/E 8 Part load optimisation 9 Waste Heat Recovery with PTI/PTO 10 Shaft generator 11 Installation of optimum sized Auxiliary Boilers 12 Reduction of HVAC Energy Consumption 13 Design for 10% lower speed 14 Design for Operational Profile

ALTERNATIVE FUEL 15 Fuel Cells for Main Propulsion 16 Use of LNG fuel

Table 1 List of ESPs suitable for construction

TARGETS

8/94

2.4.1 Short description of selected construction ESPs

Contra Rotating Propellers The contra-rotating propeller system is an energy saving device which consists of two propellers where one is positioned downstream of the other, partly or totally overlapping and rotating in the opposite direction. The contra-rotating propeller system has the hydrodynamic advantage of recovering part of the lost slip-stream rotational energy. Furthermore, because of the two propeller configuration, contra-rotating propellers possess a capability for balancing the torque reaction from the propulsion. The two propellers can be mounted on two thruster units, on an internal and external shaft, or one conventional shaft plus one contra rotating thruster. In addition, the system where the second propeller is freely rotating behind the normal propeller is known as Grim’s vane wheel.

Advantages • The propeller-induced heeling moment is compensated (negligible for

larger ships). • More power can be transmitted for a given propeller radius. • The propeller efficiency is increased leading to savings according to ITTC

(1990), Stierman (1986) and Ghassemi (2009) ranging from 2% in bulk carriers up to 12% in other vessels.

Disadvantages • The mechanical installation of coaxial contra-rotating shafts is

complicated, expensive and requires better maintenance. • The hydrodynamic gains are partially compensated by mechanical losses

in shafting.

Figure 2: Contra-rotating propellers

TARGETS

9/94

Axe-bow shape When a ship sails in waves, incident waves are reflected and broken at the ship’s bow and the resistance is accordingly increased. Hirota et al. (2005) based on model test results state that the sharpness of the bow shape above the calm waterline could reduce the added resistance. With a blunter bow shape, such as that of tankers or bulk carriers, waves are mostly reflected forward and so the resistance increases. The increase in wave resistance acting on such full-form ships with blunt bow is therefore larger than that on slender ships. For blunt-bow shaped full-form ships with smaller power engine, the speed loss is estimated to be larger than that for ships with conventional high power engine. To improve the performance in waves for ships with low power engine, the resistance increase in waves needs to be reduced. To do this, the bow should be made less blunt.

Figure 3: Axe-bow principle for reduction of wave-induced resistance

Results of a study on the effect of bow bluntness on the resistance increase indicated that the most effective way was to sharpen the bow shape above the still water level, where the wave surface is elevated and reflected. By sharpening this part, the incident wave is reflected mostly to the side, not forwards, thus reducing the wave resistance acting backward. Figure 3 indicates that “Ax-Bow” design may reduce the resistance increase in waves by 20% to 30% in almost the entire range of wavelength. This enables a 4 to 6% reduction of horsepower, or fuel consumption, in the case of sea conditions corresponding to a 20% sea margin.

TARGETS

10/94

Part load optimization

In order to meet the charter party agreement and reduce fuel oil consumption, the ship’s speed is altered to maximize economy. Moreover, the transportation chain currently does not require non-critical cargo to be transported at high speeds. This drive often results in operating the engine at a reduced load, which in turn has placed more emphasis on operational flexibility in terms of demand for reduced SFOC (Specific Fuel Oil Consumption) at part/low-load operation of the main engine. However, on two-stroke engines (camshaft or electronically controlled), reduction of the SFOC is affected by NOX regulations in order to maintain compliance with the IMO NOX Tier II demands. As it can be observed from Fig. 4, the fuel savings are up to 2 g/kWh at the part-load tuned point, while maintaining NOX compliance. It has been reported from the industry that fuel savings can reach up to 1.5% compared to similar ships without part-load optimized Main Engine. As mentioned above, the SFOC is limited by NOX regulations on two-stroke engines. In general, the NOX emissions will increase if the SFOC is reduced and vice versa. The engine is optimized close to the IMO NOX limit, which is why the NOX emissions cannot be increased. Furthermore, the new configuration of part-load has an Exhaust Gas Bypass (EGB) system. The EGB system is tailored at 6%. The Main Engine is matched with a Variable Turbine Area (VT) Turbocharger. With this method, the area of the nozzle ring is altered increasing the part-load efficiency of the T/C.

Figure 4: Effect of Part-load Optimization in Specific Fuel Oil Consumption of Main Engine (Source: MAN

Diesel)

TARGETS

11/94

Asymmetric Stern:

During a fully automated optimisation using the newly developed adjoint solver of FreSCo+ the optimisation of a bulk carrier with active propulsion yielded a asymmetric stern shape. This result evoked the success these hull forms initially created by Nönnecke in the 1970ies and 80ies.

Starting with the parametric form description of a baseline design, the optimisation was set up using the delivered power (PD) as an objective function in the adjoint equation. Computations were run in “propulsion mode” using the well-established RANS-BEM coupling algorithm in FreSCo+. The form modifications followed the predicted sensitivities on the hull depicted in the figure above. The result obtained after less than 20 iterations indicated a reduction of PD for a target speed of 14 kts by more than 2.5%. To validate the computational results, model tests were performed for the – symmetric – base line design and for the optimised asymmetric aft body in HSVA’s large towing tank. The experiments confirmed the superior performance on the asymmetric hull, indicating an even greater reduction of more than 4% for the design speed.

Although the asymmetric hull form is clearly not an option for retro-fit of existing vessels, the concept – once again – indicates a clear way forward for new ship designs with improved propulsive efficiency. The main advantage of the design is the negligible penalty on resistance. Simulations as well as the experiment indicate that the total resistance of the bulk carrier was not affected by the asymmetric stern shape. This is hardly the case for “external” energy saving devices such as fins or ducts which can also be found on newly built ships to improve propulsive efficiency. The present design approach used for the bulk carrier also maintains a vertical stem contour which limits the asymmetry to frame sections inside the hull only. This is expected to further ease the production of such vessels.

TARGETS

12/94

Figure 5: Wake distribution of the asymmetric stern

2.5 ESPs for retrofitting These ESPs generally cost less to apply and have a rather short investment return period. In addition to that, they are easier to apply without extensive alterations to the ship’s hull and machinery. These attributes make them ideal for retrofitting. The following list includes a few of the most popular commercially available ESPs at the moment and while some of them have only recently emerged (solar panels), many are common practise (hull cleaning). The reason the latter have been included is for the quantification of their contribution and comparison against the rest. The same applies to this table as the previous one which means it is neither exclusive nor exhaustive. New energy saving technologies emerge all the time and depending on the economic viability some of the ESPs presented suitable for construction can be retrofitted while most of the ESPs presented as suitable for retrofitting can and should be applied in the construction stage.

TARGETS

13/94

No. ESP HYDRODYNAMIC AIDS

1 Rudder surf bulb 2 Rudder surf fins 3 Propeller boss cap fins 4 Mewis duct 5 Propeller duct 6 Wake equalising ducts 7 Pre swirl fins 8 Pre swirl stator 9 CLT or Kappel propellers 10 Grim vane wheel

RESISTANCE MINIMISATION 11 Silicon anti-fouling paints 12 Propeller polishing 13 Hull cleaning 14 Air lubrication 15 Minimisation of resistance of appendages 16 Optimisation of bulbous bow for slow steaming

MACHINERY 17 Fuel injection slide valves 18 Electronic governors 19 Turbocharger isolation 20 Cylinder isolation 21 Replacement of incandescent bulbs with CFLS and TFLS 22 Replacement of motors with higher efficiency ones (EFF1) 23 Installation of fuel Oil Homogenisers

ALTERNATIVE FUEL 24 Solar panels for auxiliary loads 25 Kites and sails

Table 2 List of ESPs suitable for retrofitting

TARGETS

14/94

2.5.1 Short description of selected retrofitting ESPs Rudder surf bulb The goal of the application of energy saving devices in rudders is to increase the energy recovery ratio from the propeller losses since the rudder is located downstream of the propeller. There are three main sources of propeller losses: frictional, axial and rotational losses. Whenever the rudder is placed downstream of the propeller, rotational losses are recovered.

Figure 6: Propeller Forces Acting on Rudder (left), Installed Combination of Costa Bulb and Transversal Fins

on Tanker (centre) and Costa bulb (right).

There are many ways of doing this; one could be modifying the geometry of each horizontal profile of the rudder and adapting it to the velocity field. Other solutions could use devices such as Costa bulb type, or employ transversal fins. Towing tank facilities correlated the model test results for transverse fins to full scale values and state that up to 5% savings can be expected.

The Costa bulb is installed in order to:

• Reduce the hub vortex • Increase the wake fraction • Reduce the contraction and provide homogeneous axial propeller

slipstream • Reduce pressure pulse induced by propeller

Experimental facilities converge to the value of savings due to rudder bulb which is up to 2%. A correct installation of the combination of the two technologies can lead to the sum of savings.

Mewis duct The Mewis Duct is a novel power-saving device which has been developed for slower ships with full form hull shape, that allows either a significant fuel saving at a given speed or alternatively for the vessel to travel faster for a given power level. The Mewis Duct consists of two strong fixed elements mounted on the vessel: a duct positioned ahead of the propeller together with an integrated fin

TARGETS

15/94

system within. The duct straightens and accelerates the hull wake to the propeller and also produces a net ahead thrust. The fin system provides a pre-swirl to the ship wake which reduces losses in propeller slipstream, resulting in an increase in propeller thrust at given propulsive power. Both effects contribute to each other. The achievable power savings from the Mewis Duct are strongly dependent on the propeller thrust loading, from 3% for small multi-purpose ships up to 9% for large tankers and bulk carriers. The power saving is virtually independent of ship draft and speed. The Mewis Duct ideally suits both to new-buildings and to retrofit applications (e.g. on Tankers, Bulk Carriers).

Figure 7: Fluid Flow at Stern where a Mewis Duct is Present

High blocked ships have low propulsion efficiency. The reasons are the bad wake field and the high propeller loading. The water inflow has such an unfavorable characteristic that the propeller is working in bad inflow conditions. The Mewis Duct harmonizes and stabilizes the flow and generates a pre-swirl to reduce the rotational losses in the propeller slipstream. The integrated fins have a stator effect by generating a pre-swirl in counter direction of the propeller operation. This generates more thrust. The fins are asymmetrically profiled and arranged to generate a perfectly homogenous flow distribution.

Figure 8: Detailed Schematic of Mewis Duct System (Source: Becker Marine Systems)

TARGETS

16/94

All hydrodynamic aids work pretty much in similar ways, either by enhancing the flow to the propeller or minimising the cavitation etc.

Silicon-based antifouling paints Anti-fouling coatings are used to improve the speed and energy efficiency of ships by preventing organisms such as barnacles and weeds from building-up on the underwater hull surface, thus increasing the ship’s friction resistance. They may contain biocides or be biocide-free. Biocidal antifoulings’ effectiveness depends on both the biocide itself and the technology used to control the biocide release, polishing or leaching rate. All the anti-fouling coatings must be in compliance with the International Convention on the Control of Harmful Anti-Fouling Systems on Ships (AFS Convention) which restricts the use of organotin biocides in anti-fouling paints used on ships. Some of the traditional AFS-compliant paints act as biocidal anti-fouling systems and are formulated as self-polishing polymer coatings that hydrolyze and polish. Some other coatings - to a degree - wear away as the ship is propelled through the water, to expose a fresh layer of biocide, thus having an inherent added resistance by formulation.

Propeller polishing Routine in-service polishing of propeller reduces its surface roughness caused by organic growth and fouling. There is evidence that the effects of a poorly maintained propeller can decrease speed and fuel efficiency compared to that of a propeller maintaining an "A" finish on the Rupert Scale. Propeller polishing procedure can be integrated with the time spent in port of calls / waiting at anchorage, thereby ensuring that there is no loss of operational time due to the maintenance.

Taking into account the above, propeller polishing should be carried out when possible. The polishing standard should be of no less than Rupert B on any parts, confirmed by the polishing provider.

TARGETS

17/94

Figure 9: Underwater propeller polishing

Hull cleaning

Hull cleaning should be carried out based on a condition assessment basis. Therefore, in conjunction with every propeller polishing, the hull should be inspected for damage and marine growth. If there is significant growth on the hull, an immediate decision to clean the hull could be made by the Company taking into account the report / notification by the Master.

If micro-fouling is present, a hull cleaning by soft brushes should be considered for the next propeller polishing time. It should be noted that macro-fouling hull cleaning is restricted in many ports, and some ports do not allow ships with macro-fouling to enter. Furthermore, niche area fouling is coming under scrutiny and these should be inspected, and addressed if necessary, each time when propeller cleaning is under taken.

As it is evident from the following table, hull fouling can significantly increase the drag on a ship, thus reducing the speed and increasing the fuel consumption, which is usually the case for most ships approaching their dry-docking due time.

ks (m) ks+ ΔCF ΔCF(%)

1. Typical antifouling Schultz (2004) 0.00003 6 6.38677E-05 4.41% 2. Best FR 0.000016 3.2 0 0.00% 3. Best SPC 0.000019 3.8 1.63547E-05 1.16% 4. Best CDP 0.000018 3.57 1.24295E-05 0.89% 5. Light slime / Deteriorated coating 0.0001 21.7 0.000319692 18.77% 6. Heavy slime 0.0003 70.4 0.000623591 31.05% 7. Small calcareous fouling or weed 0.001 258 0.001095931 44.09% 8. Medium calcareous fouling 0.003 845 0.001618063 53.30% 9. Heavy calcareous fouling 0.01 3000 0.002095555 60.17%

Table 3 Fouling degree (Source: Schulze)

TARGETS

18/94

Turbocharger cut-out

The Turbocharger Cut-Out system is designed to lower the fuel oil consumption and improve the main engine performance during part-load operation. Turbocharger Cut-Out can be achieved in two different ways, either by installing swing gate valves on the TC air outlet and TC exhaust gas inlet or by installing blinding plates on the TC air outlet, TC exhaust gas inlet and outlet. However, installation of swing gates on engines with only two turbochargers is recommended. The obtainable load range after one turbocharger is cut-out can be found in Table 2 below.

No of T/C 1 of 2 2 of 3 3 of 4 Load Range 10-40% MCR 10-66% MCR 10-74% MCR

Table 4: Obtainable load range after one T/C is cut-out

Figure 10: Turbocharger cut out with a swing gate valve (Source: MAN B&W)

On engines with three turbochargers, one turbocharger cut-out enables operation at loads from 20% to 66% MCR, delivering:

• An expected SFOC reduction of 5 g/kWh and a 0.25 bar increase in scavenge-air pressure at 25% power,

• An expected SFOC reduction of 3 g/kWh and a 0.52 bar increase in scavenge air pressure at 50% power,

• Turbine-out temperature drops of up to 30 degrees.

Engines with four turbochargers and one turbocharger cut-out enables operation at loads from 20% to 74% MCR, delivering:

• A SFOC reduction of 6 g/kWh per 0.15 bar increase in scavenge-air pressure at 25% power

• A SFOC reduction of 5 g/kWh per 0.41 bar increase in scavenge-air pressure at 50% power

• Turbine-out temperature drop of up to 50 degrees

TARGETS

19/94

Cylinder cut-out

Another measure to increase the fuel efficiency of a Marine Diesel engine in low loads is the introduction of Cylinder cut-out system. The cylinder cut-out system (Fig. 10) to be used at RPM below 40% of MCR RPM, allows the engine to operate with only half of the cylinders, resulting in increased load on the operating cylinders with improved operating conditions for the fuel system as a result, thereby ensuring stable running conditions down to 20-25% of nominal RPM. The speed limits for the actual plant should be evaluated by the manufacturer.

Figure 11: Cylinder cut-out system

2.6 ESPs for operation These are ESPs that differ from the previous categories in that they don’t necessarily need modifications to the vessel but merely a change in the operational practices. This means that in most cases they cost little, if any, in money and – especially when applied in proper combinations – can lead to significant savings and emissions reduction. In addition, experience in energy audits onboard vessels has revealed numerous malpractices that regularly occur in ship operation that if minimised or eradicated can also help towards a more efficient vessel. The following table presents a list of recognised and available operational practises of the sort. It also comprises a notion of cost and claimed savings according to practise and claims from promoters.

TARGETS

20/94

No. ESP Cost Savings 1 Slow steaming negligible/indirect High 2 Virtual port arrival Zero High 3 Propulsion efficiency monitoring High High 4 Weather routing software High High 5 Port turn-around time Zero High 6 Ballast & trim optimization Low / Medium High 7 Speed optimization Negligible High 8 Autopilot upgrade/ adjustment Low Low 9 Optimised voyage planning Negligible High 10 Optimisation of use of fans and pumps Zero Medium 11 Optimisation of use of bow-thrusters Zero Medium 12 Efficiency control of HVAC system Zero High 13 Speed/ Power control units for electrical equipment Low Medium 14 Cargo heating and temp. control optimisation Zero High 15 Optimum lighting operation management Zero Low 16 Usage of Fuel Oil additives Low High 17 Ballast water exchange optimisation/minimisation Zero Medium 18 Even main engine / e-load operation Zero Medium 19 Elimination of voltage unbalance in motors Low Low 20 On-shore power supply (Cold Ironing) High High 21 Proper use of FO purifiers Zero Medium 22 Improved Machinery Maintenance1 High High 23 Energy Management High Medium

Table 5 List of ESPs for operation

As can be seen in the previous table most of the proposed ESPs cost nothing and promise increased savings. There are also a number of practises that could require an increased capital investment but claim high savings thus having a return period of less than 3 years in the worst case. An analysis carried out in a previous work package and can be found in D4.3 resulted in the following distribution for investment return period.

1Minimization of Air System Leakages, Proper insulation of steam distribution network, Electrical insulation of Electric Network and overhauling of M/E and A/Es at specified by the manufacturer intervals.

TARGETS

21/94

Figure 12: Investment return period (in months)

This distribution includes all the ESPs presented so far from all construction, retrofitting and operational chapters but is indicative of the time it takes for the investment to return profit. Half the ESPs require less than 12 months while 90% takes less than 3 years.

2.6.1 Short description of selected operational ESPs

Virtual arrival

It is inherently wasteful for a vessel to steam at full speed to a port where delays to cargo handling have been identified. By reducing speed to reach the destination at a mutually agreed arrival time, the vessel can avoid spending time at anchor awaiting berth, tank space or cargo availability. Emissions could thus be reduced, congestion could be avoided and safety in port areas could be improved. The potential energy savings for just-in-time arrival is assessed at 1-5%. The highest potential savings would be expected where economic considerations (incentives from contractual agreement) favour inefficient operational arrival. The Virtual Arrival concept is one example of the coordination between the owner and charterer having been further developed. The following pre-requisites apply:

• A known delay at the discharge port; • A mutual agreement between the company and the charterer; other

parties may be involved in the decision making process (e.g. Terminals, cargo receivers and other parties sharing commercial interests);

TARGETS

22/94

• An agreed charter party clause that establishes the terms for implementing virtual arrival;

• An agreement on how to calculate and report the performance of the vessel;

• An agreement on how to assign benefits between the involved parties.

A typical operational profile of a Capesize Bulk Carrier using voyage data for one year is given in the figure below which provides the time percentage for the various ship’s operating modes (sea passage - ballast & laden conditions, anchored/drifting, loading, discharging, pilotage & alongside).

Figure 13: Typical Vessel Operational Profile by Mode

Anchorage-drifting time is directly related to “Just in time” arrival, and can be attributed to the congestion at the loading/discharging terminals, delays due to weather conditions, etc.

D/G Engine Load Optimization and Electric Load Demand Minimization

Low D/G loads (below 40%) have an adverse effect to the operation of the engine (particularly the FO system and cylinders) leading to increased maintenance costs and accelerated wear of engine components. Due to these reasons it is prudent to exercise efficient load management, with the aim to minimize the number of running generators and maximize their load factor, when possible and safety permitting. The total electricity demand will be the same, but less operating engines at higher load, i.e. lower SFOC, translates to reduced fuel consumption. The following table provides an example of a generic guide on the number of D/Gs to be in operation depending on the vessel’s operational mode.

TARGETS

23/94

Operational mode Number of D/Gs in operation At sea One (1) At anchor One (1) Stand-by Two (2) Discharge operation Two (2) Loading Operation One (1)/

Two (2) when de-ballasting Ballast Water Exchange, Tank Cleaning, Purging Two (2)

Table 6 Guidance on the Number of D/Gs in Operation

Optimised use of pumps

A simple example of efficient electric load management is the operation of the steering gear pumps. According to the ship’s Electric Load Analysis booklet, steering gear hydraulic pumps are not required to run whilst the vessel is in port. Reduction of steering gear hydraulic pumps running hours, via switch-off when the vessel is in port, is suggested. The instruction to the deck officers would be: Stop steering pump in port after “finished with engine”. If the subject pumps are not switched off during port stay then the expected increase in the fuel consumption can be up to 10 MT per year for a Capesize Bulk Carrier. Another example is the M/E LO and the Camshaft LO Pumps, which can be switched off in port. Many terminals require the M/E to be ready on short notice so it might not be possible to implement the above strategy at all times, but same should be considered when possible. The implementation of this instruction is up to the Chief Engineer‘s discretion.

E/R Fans operation management

The E/R fans role is to supply air for combustion to the diesel engines and to ensure adequate air circulation in the E/R. The E/R fans should be operated taking into account the E/R air balance study and the actual temperatures in the E/R. Under mild weather conditions it should be possible to stop one or two of the E/R fans when the vessel is not sailing, i.e. the M/E is stopped. The E/R fans role is to supply air for combustion to the diesel engines and to ensure adequate air circulation in the E/R. The E/R fans should be operated taking into account the E/R air balance study and the actual temperatures in the E/R. Under mild weather conditions it should be possible to stop one or two of the E/R fans when the vessel is not sailing, i.e. the M/E is stopped. Implementation of the above strategy is expected to additionally lead to reduced maintenance costs for the E/R fans due to lower operating hours.

TARGETS

24/94

3 Technologies addressed and developed in the project

3.1 Improved CFD methods for resistance prediction

3.1.1 Overview In task T1.1.1, associated with deliverable 1.1, HSVA and TUHH significantly improved the capabilities of the RANS solver adFreSCo+. The developments focussed on the new adjoint module as well as the improvement of the prediction of multi-phase-flows. The developments of task T1.1.1 allow the partners to efficiently assess given ships geometries for their hydrodynamic optimisation potential. The new adjoint module gives considerable insight in the optimisation potential of analysed hull geometries. Industrial objective functions for wake field assessment were integrated to the adjoint formulation. The calculated sensitivities can either be applied manually by visually interpreting the results of the adjoint analysis or can be mapped to the parametric geometry definition of a CAE-framework. The latter is based on the newly developed mapping interface that translates the discrete sensitivities to the continuous CAD parameter space. The applicability of the adjoint solver in conjunction with active propulsion was demonstrated. The adjoint solver proves to allow stable simulations at full-scale Reynolds-numbers of up to 1E9. The volume-of-fluid module was improved to allow for more efficient and accurate simulations. A sub-cycling technique that allows specific time-step was introduced. A speed-up of factor 30 was shown for the wave resistance prediction of the well-known KCS container ship test case. Further improvements focussed on the allowable courant numbers that restrict the size of the time-step. The improvements give a considerable speed-up of the simulation as well as an improved resolution of the predicted free surface regime. Besides algorithmic improvements the efficiency of the RANS free-surface (VoF) calculations was improved by coupling the RANS solver with the HSVA free-surface panel code v-Shallo. The source field of the panel-code solution is employed to initialise the free surface for the RANS-VoF calculation as well as the boundary layer flow. Furthermore, the trim determined by the panel-code is considered. These changes lead to a speed-up of 50-70% for the RANS-VoF calculation.

TARGETS

25/94

3.1.2 Results

Optimisation of flow to propeller

The following is an example of the capability of new developed adjoint solver adFreSCo+ in calculating the sensitivity of the aft. end of a container vessel to geometric changes that influence the wake to the propeller. The new adjoint solver is applicable to both model as well as full-scale Reynolds numbers. The results obtained at model scale, as shown in figures 13 and 14, reveal major differences in the wake field directly upstream of the propeller disk. For the unpropelled case a significant velocity defect appears below the propeller shaft which is caused by the pronounced knee at the aft/lower part of the skeg. Accordingly, the sensitivities indicate that a smoother transition from the keel line up to the propeller shaft would enhance the wake field quality.

Figure 14: Results without propulsion. Contourplot of sensitivities (left) at a Reynolds number of 1E7 (red

[blue] indicates a reduction [an increase] of displacement). Corresponding wake field upstream of the propeller disk (right).

Figure 15: Results with propulsion. Contourplot of sensitivities (left) at a Reynolds number of 1E7 (red [blue] indicates a reduction [an increase] of displacement). Corresponding wake field upstream of the

propeller disk (right).

TARGETS

26/94

Figure 14 shows the results of a simulation that considers the influence of the propeller. The velocity defect below the shaft line is significantly decreased by the momentum of the propeller flow. The information given by the sensitivities is similar to the unpropelled case, but at a much lower scale. The results show, that the adjoint method can successfully be applied in conjunction with a propeller model that imposes the propeller flow by distributed body forces.

Improvements of Free Surface (VoF) Computations

Ship hull form optimisations depend largely on accurate predictions of the “wave resistance problem”, i.e. the flow around the ship’s hull at the free surface. Improving the hull form to reduce the ship’s resistance is one of the prime objectives of all model basin work. The following figures show how an improved and faster calculation method results in accurate results at least 3.5 times faster. Figure 15 shows the resulting wave pattern calculated with the original scheme while figure 16 shows the corresponding result calculated with the modified HRIC scheme. It can clearly be seen, that a significant part of the ship’s wave features are lost in figure 15 due to numerical diffusion, while in figure 16 a very clear and detailed wave pattern is present.

Figure 16: Free surface elevation calculated with original scheme and parameters.

Figure 17: Free surface elevation calculated with original scheme and parameters.

TARGETS

27/94

3.2 Improved hull forms The following summarise the application of the developed tools presented earlier to a number of different ships being representative for typical cargo vessels, container ships as well as bulk carriers or car transporters. Different key aspects of ship design affecting the resistance are treated. These vary from global considerations of the overall ship concept, e.g. the choice of mono-hull or multi-hull designs to individual hull form optimisations and the consideration of typical details such as bulbous bows and appendages which will have a considerable influence on the final hydrodynamic performance of a ship. The three main elements used in this form of optimisation are in general terms:

• Geometry definition and modification tools; • Analysis – CFD tools; • Optimisation tools.

3.2.1 Bulbous Bow Optimisation for Slow Steaming A large container vessel with a capacity of 8,500 TEU was designed for operation at speeds well in excess of 24 kts a few years ago. Today, transit speeds have dropped considerably and due to lower cargo volumes the full design draft is often not reached. This calls for a thorough analysis of the initial vessel design at lower speeds and possibly design modifications in order to optimise the ship for new target operational conditions. To improve the resistance characteristic at lower speeds a modification of the bulb has been considered. The bulb which was designed for high draft and high speed (> 24 kts) conditions clearly performs sub-optimal for low speed / draft combinations. The general tendency will be towards a shorter, smaller bulb to improve the situation. A complete parametric study of different influence parameters has been carried out. The main options that have been compared are:

• Shortening of the bulb • Reducing the bulb volume

The following figure shows a schematic of the modification. The blue outline shows the original hull.

TARGETS

28/94

Figure 18: Bulb modification. The blue lines show the original hull.

The predictions have been performed for a draft of 11.5m and a draft of 13 m. The following figures show equivalent results for the latter case. It can be seen that version Modi_6 shows the best resistance characteristic saving about 14% total resistance in average over the speed range considered.

Figure 19: Total resistance - comparison of different bulb geometries at T = 13.m

The following exemplary plots show the variation of the pressure distribution on the bow for the original and the optimal variant, each at the same speed and draft condition.

TARGETS

29/94

Original

Modi_6

Figure 20: Hull Pressure and wave elevation at the bow, T = 11.5m, v = 12 kts

3.2.2 Multihulls The reason(s) why the multi hull concept so rarely found in larger cargo carrying commercial vessels have been studied. The added complexity of the build is clearly an issue which affects overall cost of a new ship. This aside, studies were focused on the resistance alone and hence highlighted the effect on operating cost. In the following, a comparison is presented between a monohull and a catamaran configuration with the same displacement.

Figure 21: Comparison of wave pattern - Catamaran (left) and Monohull (right)

TARGETS

30/94

The non-dimensional resistance coefficients cF (frictional resistance, ITTC line), cR (residual or wave making resistance) and cT (total resistance) are commonly used to compared different ship types and sizes. The following figure shows these resistance coefficients comparing the monohull with the catamaran.

Figure 22: Resistance coefficients for a catamaran, monohull and trimaran with equivalent displacement

While the comparison of dimensionless resistance coefficients shows a clear advantage of the multi-hull configuration, for a final assessment of the overall resistance of the two ship types not only the coefficients but also the size and especially the wetted surface need to be taken into account. This however has increased significantly from monohull to catamaran by a factor of almost 1.7. The following figure shows the results for the total resistance of 3 ship types. While the resistance coefficients for the catamaran are clearly better than those of the other two, the complete resistance however shows a different picture, due to the increased wetted surface. From this it is clear that in terms of overall resistance the catamaran is only superior for the highest speed corresponding to a Froude Number of 0.31.

TARGETS

31/94

Figure 23: Total Resistance - Comparison of Trimaran, Monohull and Catamaran of equivalent

displacement

3.2.3 Aft body optimisation

Case 1: Improvement of wake field

An extensive investigation was carried out in order to determine how to reduce the delivered power, PD, without sacrificing other aspects of ship performance such as sea-keeping, manoeuvring, and cargo load capacity. The propeller performance is the focus of this investigation. The same adjoint solver that was used in previous has been used for this study, providing a guideline for reducing the power requirement, utilising a goal function that combines the hull resistance and the propeller inflow homogeneity. These two flow properties have a strong influence on propeller performance. The adjoint method generates a new result variable, called the sensitivity values. These sensitivity values show where a hullform change can have the greatest impact on the associated goal function. For this set of computations on a capesize bulk carrier, the regions of highest sensitivity to form change according to the adjoint computation are shown in the following figure.

TARGETS

32/94

Figure 24: Adjoint Sensitivities, Profile View (upper) and aft view (lower)

A small number of control points are used to modify larger regions. The mathematical basis ensures a smooth transition from the modified geometry into the unmodified geometry, thus avoiding discontinuities. The optimisation results are shown in the figure 25 following.

Figure 25: Optimisation Results for Propeller Performance

Figure 26: Axial Component of Propeller Wake, Baseline (left), Improved Skeg (right)

The needed delivered power reached a minimum at variant 5, demonstrating an improvement of approximately 1.3%. The improvement in the wake flow is shown in figure 26, where the increased homogeneity is visible.

TARGETS

33/94

Case 2: Transom wedge

The example shown in this section focuses on the transom stern wedge variations on a 110.000 DWT Aframax tanker vessel. Transom stern aft form is quite common on modern hull forms. Stern wedge is not normally employed for conventional ships and is usually preferred for high speed hull forms. The investigation here involves the resistance effects of 3 transom wedge lengths and 3 transom stern. The variants are showed in the following figures. Calculations showed that the stern wedge mostly increases resistance in the operational envelope of such a vessel, thus is not an effective measure for increasing efficiency. There is however a small benefit of approximately 1% to be gained as shown in figure 29. For faster RoRo / RoPAX ships improvements of up to 7% have been reported recently.

Figure 27: Original aft hullform

Figure 28: Modified aft hullforms

TARGETS

34/94

Figure 29: Effect of wedge height on resistance

3.2.4 Optimised appendages A wake homogeneity enhancing duct configuration by means of a duct positioned upstream of the propeller has been investigated. The underlying concept is to accelerate the axial flow within the duct at the cost of a slight deceleration outside of the duct (Schneekluth et al. 1998). As single screw vessels are usually afflicted with a region of rather low axial velocity just around the 12 o’clock position, this interplay of acceleration / deceleration can be advantageous in terms of circumferential homogeneity of the axial propeller inflow. The axial propeller inflow has a strong impact of the fluctuation of the blade load and thus on the occurrence of cavitation and pressure pulses of high magnitude. Several duct configurations where created by manually adapting the shape parameters of the CAD model. The initial study is supplemented by an adjoint shape optimisation study to further enhance the duct configuration. To allow for an efficient evaluation of different geometric configurations, a fully parametric CAD model is employed, the aft lines plan of the considered ship is shown in the following figure.

Figure 30: Aft lines plan of studied vessel with indication of duct and propeller

TARGETS

35/94

The contribution shows, how retrofitted propulsion augmenting devices can be employed to increase the wake homogeneity in the critical 12 o’clock position of the propeller. This was achieved by carefully positioning the device in the aftship flow. The presented study found a configuration that enhances the wake quality while generating thrust. For such retrofit solutions it appears beneficial to enhance the initial configuration by optimizing the retrofitted shape with respect to the longitudinal force. Here the adjoint optimisation study resulted in an increase of 46% of the longitudinal thrust of the wake equalising duct while maintaining the homogeneity of the wake field.

3.3 Improvement of propulsion A fraction in excess of 10% of the available energy is thought to be lost in propulsion. Be it frictional losses, poor quality of flow to the propeller or low performance of the propeller itself, it has to be identified, measured and reduced.in the following sections, a number of methods to reduce losses in propulsion will be presented, from unconventional propellers to devices that improve flow to the propeller.

3.3.1 Design of optimum screw propeller A successful ship design in terms of ship powering demands propulsion devices designed to give maximum efficiency and to absorb a low shaft power at an adequate revolution number with low hull pressures, noise, cavitation erosion and vibration. In the basic design stage, the main particulars of propellers (mainly pitch, shaft speed and blade area) are selected in iterative manner for the optimum propulsive efficiency and minimum risk of cavitation for the maximum propeller diameter. In the detailed design stage, mathematical methods are required to optimize the propeller characteristics, cavitation and vibration. Using unconventional propellers one can achieve remarkable savings, thus an analysis of the alternatives is presented here, alongside case studies. Specifically, three systematic propellers, B-series, Enhanced Meridian Series and Tip Raked series, are summarized.

B-Series

The Wageningen B-screw series is a general purpose, fixed pitch, non-ducted propeller series which is used extensively for design and analysis purpose. The origin of the B-series goes back to the development A-series, in which the blade sections were “modern” airfoil profiles having a blade contour with narrow blade tips. This blade contour, in combination with a constant radial pitch

TARGETS

36/94

distribution, made them susceptible for cavitation. This has given rise to the development of B-series with a wider blade tip and adjusted contour to improve the cavitation behaviour. Over the years, this model series has been expanded to provide a comprehensive fixed pitch, non-ducted propeller series. This includes 2-, 3-, 4-, 5-, 6- and 7-bladed series with various blade area ratios.

Enhanced meridian series

The Meridian model propeller series is a unique standard series based solely on practical propeller designs for standardised variations in pitch (P) and blade area ratio (BAR). All the parent propellers were designed to work in a uniform wake stream, the blade pitch distribution and the final section shapes being obtained from the „optimum centreline camber design‟ process. The upgraded Meridian series originated from the previous version of the Meridian series developed at the Stone Manganese Marine (SMM) Ltd and the University of Newcastle upon Tyne. The upgraded series now includes 3-, 4-, 5- and 6-bladed designs with various blade area ratios (EAR = 0.45, 0.65, 0.85, 1.05) and pitch-diameter ratios (P/D = 0.4, 0.6, 0.8, 1.0, 1.2).

Tip raked series

Amongst many unconventional types of propulsors, perhaps one of the simplest and most effective is the so-called “Tip-raked” (or tip plated) propeller where the blades are bent at the tip to improve the propeller’s efficiency by reducing the axial losses in its slipstream. The tip rake propeller blade is described by a 90° bending of the blades from the vertical plane to the horizontal at the outer radii (see figure 31).

3.3.2 Fixed swirl-propeller devices Both post-swirl (downstream stator) and pre-swirl (upstream stator) devices are designed such that the tangential velocities, which they induce in the slipstream, cancel those induced by the propeller (or rotor). The downstream stator has a negligible effect on the propeller forces but, for appropriate propulsor loading, the stator produces a net positive thrust and the propulsor efficiency becomes greater than that of the equivalent conventional propeller. On the other hand, the upstream stator produces a net negative Figure 31: Tip-Raked propeller blade

TARGETS

37/94

thrust but modifies the flow to the propeller in such way that the propeller thrust is increased and, again in the right conditions, the propeller efficiency is increased. As far as the selection of the propeller/stator arrangement is concerned, a series of design case studies was investigated for the upstream and downstream stator/propeller combinations. This investigation indicated consistently more gain in efficiency for the downstream stator/propeller combination than the upstream stator/propeller. The results are shown in figure 32 while an arrangement of a post-swirl stator for a large bulk carrier is shown in figure 33.

Figure 32: Comparative gain in efficiencies for a downstream (post) stator/propeller and upstream (pre)

stator/propeller arrangement including the effect of stator blade number

Figure 33: Post-swirl stator arrangement for large bulk carrier

TARGETS

38/94

3.3.3 Contra-rotating propellers A system of contra rotating propellers (usually a pair) is supposed to recover rotational energy that would be left behind a single propeller to 100%, if no additional steady devices would serve to recover them. As stationary pre- and post-swirl stators can be introduced for a recapture of rotational energy one may ask whether a contra-rotating unit of two propellers shows an improved performance. Thus a few advantages of the CRP system are summarized in the following:

1. In the CRP case, the optimum diameter for the recovery unit (i.e. the diameter of the aft propeller) seems to be well defined by the slipstream contraction. For the fixed stators the optimum diameter may be harder to identify. It may be larger than the propeller diameter, even if the stator is operating behind the propeller.

2. In a standard CRP design procedure the system is optimized and interactions are considered bilateral, not one-directional. For a stator/propeller system is seems to be standard to accept the stand alone propeller design and to find the optimum stator subsequently.

3. In the CRP case the hub vortex is eliminated conveniently and drag effects due to the hub vortex are of no concern (contrary to a standard single propeller arrangement).

The biggest downside is the mechanical complexion of the system because of the two shafts that are needed to operate one within the other, with the added downside of increased losses which may limit the predicted hydrodynamic gains.

3.3.4 The boundary-layer alignment device (BLAD) The flow conditions in the wake behind a ship play a crucial role for the propulsive efficiency. The shape of the aft ship determines the inflow into the propeller. Especially the bulky hull forms of tankers and bulk carriers suffer from massive losses of axial velocity above the propeller shaft. Hence, a novel type of hull appendage for improved propulsion is presented, namely BLAD (boundary-layer alignment device). The purpose of the BLAD is to deflect the outer streamlines in the aft-ship flow field towards the hull surface so as to accelerate the wake flow and reduce the boundary-layer thicknesses. In this way an enhanced, more homogeneous flow through the propeller shall be achieved. In addition, the BLAD design intends to create a swirl against the propeller rotation so that the propeller can generate extra thrust at a given rate of revolutions. In contrast to the well-established

TARGETS

39/94

propulsion-improving devices (e.g. stator fins and propeller ducts), which are usually mounted close to the propeller, the BLAD is applied further upstream of the propeller (see Figure 34) such that a large portion of the wake field can be modified. This approach allows for a more drastic change of the propeller inflow and possibly for a larger gain in energy saving. However, it may turn out that the propeller needs to be adapted to the new flow conditions, rendering the BLAD concept a more expensive retrofit alternative as compared to the established propulsion-improving devices. The upstream placement of the BLAD is also advantageous in that the device is not exposed to the propeller-induced unsteady flow field, i.e. there are no dynamic loads acting on the BLAD structure.

Figure 34: The BLAD concept. Close-up view of the starboard deflector and position of the deflectors.

The BLAD consists of two flow deflectors mounted on the port and starboard sides of the ship hull (Figure 34). They feature a NACA wing with integrated struts and are comparable to the propeller blades in size. The BLAD appendages are placed approximately two propeller diameters upstream of the propeller plane at zero sweep in order to avoid any cross-stream flow. In the first step, the wings and struts are carefully aligned to the local streamlines, leading to a twisted geometry. Subsequently, the wings are tilted around the leading edge so as to deflect the streamlines towards the hull surface. The angle-of-attack is adjusted until a significant improvement of the wake fraction and the wake uniformity is reached. In the last step, the starboard BLAD deflector is cambered and shifted upward. This asymmetric arrangement creates a desired swirl in the propeller inflow.

TARGETS

40/94

The findings indicate that the propulsive efficiency can be considerably increased by the BLAD concept, rendering it an interesting and promising device. However, it also turns out that a propeller modification/ re-design is necessary in order to arrive at significant power savings (up to 7.4%), whereas the BLAD alone only yields a 2% power reduction. Hence, the major strength of the BLAD lies in its ability to create the necessary flow conditions for more powerful propeller designs. The need of a modified/new propeller possibly renders the BLAD concept more expensive than existent retrofit options, although the BLAD deflectors themselves are expected to be low-cost thanks to their simplicity.

3.4 Savings through operation – Effect of trim and appendages Continuing the CFD analyses of the original 8,500 TEU container vessel provided for the studies, an investigation was carried out covering the resistance and power prediction when considering appendages and trim conditions. Firstly, the effects of appendages are examined in order to determine the contribution of bilge keels and thruster openings on the resistance and powering prediction. Secondly, the investigation covers the added resistance for off-design loading conditions. Since the former are inevitably connected, they must be accounted for concurrently. An example of the appendages considered in the calculations is shown in the following figure.

Figure 35: Container Ship: Location of Appendages

Lately, the performance prediction at alternative or “off-design” conditions has become more important to ship operators. Historically, only the even keel loading conditions were considered during the design process. Only minor consideration was given to the off-design loading conditions, such as different mean drafts, or by-the-bow or by-the-stern static trim conditions. Without such information, ship operators may unknowingly cause considerable additional fuel costs, simply by sailing at sub-optimal trim conditions. In some instances, these additional costs may be avoided by shifting the load or ballast

TARGETS

41/94

to a more efficient trim condition. The following table shows the conditions that were investigated.

Drafts, T [m] 11.5 13.0

Trim Conditions 2 [m] bow up Even keel 3 [m] bow down

Speed, V [kts] 12 17 21

Fn 0.105 0.149 0.184

The computations all use adFreSCo+, relying on the free-surface, 6-Degree-of-Freedom (6-DoF) dynamic response capabilities to provide the resistance values and datasets describing the flow around the ship hull, as described earlier. In order to have a common ground for comparison, all resistance values are given in effective power PE [kW] values. The appendage geometries are meshed and computed within the adFreSCo+ computation. The results are shown in the following figures.

Figure 36: Effects of Trim on powering, Mean Draft 11.5 m

TARGETS

42/94

For the mean draft value of 11.5m and at the lowest speed, 12 knots, both bow up and bow down conditions exhibit lower powering than the even keel condition with 8.7% and 5.2% reductions, respectively. At 17 knots, the bow down condition performs even better (9.7% reduction), while the bow up condition exhibits higher powering requirements of 6.4% more than the even keel condition. Finally, the highest speed of 21 knots exhibits an increase in power required for both bow up and bow down loading conditions at 5.3% and 1.5%, respectively.

Figure 37: Effects of Trim on powering, Mean Draft 13.0 m

Results for this draft are somewhat different. The behaviour is much more linear with the bow down condition demonstrating a decrease in powering needs at lower speeds, which increases at higher speeds. Bow up requires more power at lower speeds while showing no effect at the highest. Overall, powering demand due to trim optimisation can be reduced by as much as 6.2% at lower speeds.

3.5 Assessment of Energy Savings through Reduced Resistance

8,500 TEU Container Vessel: Bulb Modification

A speed of 17 knots and a mean draft of 11.5m (slow steaming, low payload) have been considered for the 8,500 TEU container vessel. The propulsive efficiency is assumed to be 0.77. The daily savings of heavy fuel oil (HFO) and operational costs due to the bulbous-bow adaptation to slow-steaming conditions are presented in table 7. The data reveal that fuel-cost savings of 1 million US dollars can be achieved after approx. 202 days at sea.

TARGETS

43/94

8,500 TEU CONTAINER VESSEL, speed 17 kts, HFO combustion Propulsion power savings (bow modification) 1948.1 kW Energy savings 46754.4 kWh/day Specific HFO consumption 176.1 g/kWh HFO savings 8.23 t/day HFO-specific CO2 emission 3.1144 tCO2/tHFO CO2 savings 25.64 t/day HFO price (source: Bunkerworld) 600 $/t Financial benefit 4940 $/day Table 7: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the adaptation of

the bulbous bow of the 8,500 TEU container vessel to slow-steaming conditions

The operational condition chosen here is considered an average of the intended operational profile of the ship. In fact even larger savings would have been obtained if the ship would sail at even slower speeds between 12 kts and 17 kts as indicated in the analysis in the previous.

8,500 TEU Container Vessel: Operational Optimisation

The savings of energy, emissions and costs by trimming the container vessel bow-wards (3m bow down) are estimated for the same conditions (speed 17 knots, draft 11.5m, propulsive efficiency 0.77). The pre-trim leads to a propulsive power reduction by 965 kW with respect to an even-keel floating position. This amounts to fuel savings of some 4 tons and a financial benefit of nearly 2500 US dollars per day (table 8). 8,500 TEU CONTAINER VESSEL, speed 17 kts, HFO combustion Propulsion power savings (bow-wards trim) 965.0 kW Energy savings 23160.5 kWh/day Specific HFO consumption 176.1 g/kWh HFO savings 4.08 t/day HFO-specific CO2 emission 3.1144 tCO2/tHFO CO2 savings 12.70 t/day HFO price (source: Bunkerworld) 600 $/t Financial benefit 2447 $/day Table 8: Daily savings of energy, fuel oil, CO2 emissions and operational costs through pre-trimming the

8,500 TEU container vessel (3m bow down)

TARGETS

44/94

Panmax Container Vessel – Bulb Modification

In order to estimate the daily savings of energy, emissions and operational costs thanks to the bulbous bow modifications it is assumed that the vessel sails at an average speed of 21 knots with her engine running on heavy fuel oil (HFO). Under these conditions, about half a ton of HFO and 1.5 tons of CO2 emissions can be saved daily (table 9). For the TARGETS voyage data provided by the ship operator (sea passage time approx. 50 days, 22 ports of call; cf. Deliverable 4.2), the financial benefit amounts to nearly 15000 US dollars. CAP SAN NICOLAS, speed 21 kts, HFO combustion Propulsion power savings 113.8 kW Energy savings 2731.2 kWh/day Specific HFO consumption (shop trial) 176.1 g/kWh HFO savings 0.48 t/day HFO-specific CO2 emission 3.1144 tCO2/tHFO CO2 savings 1.50 t/day HFO price (source: Bunkerworld) 600 $/t Financial benefit 288.6 $/day Typical sea passage time per voyage 49.4 days (1185.4 h) Accumulated financial benefit per voyage 14254 $ Table 9: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the modification

of CAP SAN NICOLAS’ bulbous bow and accumulated savings for the TARGETS voyage data

When comparing the results of this case study with those obtained in the first case study for the larger container ship we must note that i) the optimisation performed here was intended for the design point for which the – existing – vessel is expected to be sufficiently pre-optimised already and ii) the optimisation method chosen, i.e. the adjoint approach is presently not fully adapted to the problem as the free surface influence needs to be taken into account through an iteration with external codes. Once this is fully implemented into adFreSCo+ larger improvements are expected.

Capesize Bulk Carrier – Aftship Modification

The daily savings due to the aftbody hullform modification for a bulk carrier are compiled in table 10. An average speed of 14.25 knots and a mean draft of 17.5m are assumed. For the TARGETS voyage data provided by the ship operator (120 days at sea, 9 ports of call; cf. Deliverable 4.2) we obtain a financial benefit of 41566 US dollars.

TARGETS

45/94

STAR AURORA, speed 14.25 kts, HFO combustion Propulsion power savings 136.1 kW Energy savings 3266.4 kWh/day Specific HFO consumption (shop trial) 176.6 g/kWh HFO savings 0.58 t/day HFO-specific CO2 emission 3.1144 tCO2/tHFO CO2 savings 1.80 t/day HFO price (source: Bunkerworld) 600 $/t Financial benefit 346.1 $/day Sea passage time for typical voyage 120.1 days (2881.3 h) Financial benefit per voyage 41566 $

Table 10: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the aftship modification for a bulk carrier and accumulated savings for the TARGETS voyage data

Small Bulk Carrier – Schneekluth Nozzle

It is assumed that the small bulk carrier sails at a speed of 14 knots and a draft of 9.5m. The propulsive efficiency is taken to be constant at 0.8. The original Schneekluth duct leads to a propulsive power reduction by 60.3 kW, while the savings through the optimised duct amount to 75.5 kW (table 11). As a result, a third of a ton of HFO worth 200 US dollars can be saved every day. SMALL BULKER, speed 14 kts, HFO combustion Prop. power savings (duct version 4, optimised) 75.5 kW Energy savings 1811.1 kWh/day Specific HFO consumption (shop trial) 176.6 g/kWh HFO savings 0.32 t/day HFO-specific CO2 emission 3.1144 tCO2/tHFO CO2 savings 1.00 t/day HFO price (source: Bunkerworld) 600 $/t Financial benefit 191.9 $/day

Table 11: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the fitting of a Schneekluth nozzle for a small bulk carrier

3.6 Viscous resistance Resistance reduction through optimization of hull form has substantial influence on the CO2 emissions of the ship. Wave resistance of ships has been

TARGETS

46/94

investigated thoroughly for the conventional ship forms through mathematical optimization studies, and drag reduction measures such as bulbous bow, entrance angle reduction have been introduced almost for all new ship designs. However, little improvement on the frictional resistance component was achieved up to now. As new CO2 reduction measures such as EEDI necessitates the slow speed steaming for the compliance of the regulation until new technology is introduced. Hence reduction of frictional resistance is a prime concern for ship hydrodynamics. The techniques for frictional drag reduction have been an important research topic for both aerodynamic and hydrodynamic research over 40 years.

3.6.1 Surface roughness due to biofouling Biofouling is known as the attachment of unwanted micro-organisms, animals and plants including diatoms and bacteria (slime) on immersed surfaces. There are more than 4000 species of fouling currently identified which makes the development of new environmentally benign antifouling systems a very challenging task. Biofouling can be broken down to two categories namely micro-fouling which is mainly slime and macrofouling which is animal and weed fouling, please refer to Figure 38. Generally it is known that fouling growth depends on how long the ship stays idle (either on anchorage or in port) and its cruising speed in water while sailing. However the nature of water e.g. temperature, pH, dissolved salt and oxygen concentration are affecting the fouling growth too. It is widely accepted that the presence of different molecules and organisms in the film influences the settlement of subsequent organisms.

Figure 38: Main types of fouling organisms

Increased roughness of the hull and/or the propeller can cause increase in fuel consumption. The condition of coating can also affect the roughness of hull, which can be influenced by cracking, detachment on hull coating and touch up

TARGETS

47/94

repairs of coating, Willsher, J. (2007). The application also plays a significant role as shown in Candries (2000) i.e. spraying always give a smoother texture hence lower skin friction than when the application is done using a roller. Biofouling significantly influences hull roughness hence skin friction. For the effects of biofouling, such as slime fouling, weed fouling and barnacles, refer to the following table.

Types of biofouling Increase in drag resistance Slime up to 1-2% Weed up to 10%

Barnacles up to 40% Table 12: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the adaptation of

the bulbous bow of the 8,500 TEU container vessel to slow-steaming conditions

The surface roughness affects the powering performance of a ship. From ship trial, ship resistance vary 3% and friction resistance 5% by different paint finish. Fairing the plate edges reduced the resistance by 3% and allowing the hull to foul resulted in the increases of over 30% in the ship resistance corresponding to about a 50% increase in the frictional resistance. The main coating types used nowadays are biocidal coatings such as Tin free SPCs and CDPs and non biocidal i.e. foul release systems. The following table shows the results of the analysis carried out to determine the effect of hull roughness to skin friction.

ks (m) ks+ ΔC

F ΔC

F(%)

1. Typical antifouling Schultz (2004) 0.00003 6 6.38677E-05 4.41% 2. Best FR 0.000016 3.2 0 0.00% 3. Best SPC 0.000019 3.8 1.63547E-05 1.16% 4. Best CDP 0.000018 3.57 1.24295E-05 0.89% 5. Light slime / Deteriorated coating 0.0001 21.7 0.000319692 18.77% 6. Heavy slime 0.0003 70.4 0.000623591 31.05% 7. Small calcareous fouling or weed 0.001 258 0.001095931 44.09% 8. Medium calcareous fouling 0.003 845 0.001618063 53.30% 9. Heavy calcareous fouling 0.01 3000 0.002095555 60.17%

Table 13: Effect on skin friction as compared with the hydrodynamically smooth hull

The models developed were applied on a test case of a VLCC tanker vessel by means of CFD in order to determine the effect for different Reynolds numbers, i.e. speeds. The following figure demonstrates the computational domain and grid for the aforementioned vessel.

TARGETS

48/94

Figure 39: Computational domain and grid setup for the test vessel

The results shown in figure 40 indicate a marked increase of friction, and correspondingly, total drag between smooth hull and hRM = 200µm. Further increase of resistance can be observed at further increasing roughness, albeit at a less steep slope. Interestingly, the ITTC’57 friction line, having a base roughness of 150µm, shows very good correlation with computational results at smooth surface (-3.1%). Naturally, the coefficients according to the friction line are much lower than the computed ones at higher roughness (up to -46.9%). After applying the roughness correlation as advised by ITTC’78, the results differ significantly (-27.7% to +48.8%). Starting at a lower value for a smooth surface, the two curves exhibit a cross-over at rHM ≈ 50µm, with the ITTC-line increasing to much larger values than the computed results. Nevertheless, both the computed results and the ITTC-line including roughness correction exhibit the characteristic “knuckle” at hRM between 100 and 200µm.

TARGETS

49/94

Figure 40: Normalised resistance components as function of surface roughness (computed and ITTC)

Figure 41 depicts the friction coefficient distribution on the hull for smooth surface and hR = 500µm.

Figure 41: Friction coefficient on smooth (upper) and rough (lower) hull

The application of appropriate antifouling coatings and use of marine growth protection systems or even an increase of the frequency of dry-docking can help eliminate the fouling problem and result in reduced consumption and thus emissions. The DEM model developed in the course of the project utilises the results of this section to lead to optimised operation.

3.6.2 Patterned surfaces Some frictional drag reduction attempts imitate the most efficient creatures of nature. Dolphins, sharks (figure 42) and whales are examples of such living natural designs. By studying and understanding nature’s hydrodynamic

TARGETS

50/94

mechanisms, strategies to reduce the frictional drag have been developed but not applied successfully up to now.

Figure 42: Photo of shark skin surface (scale bar 100 μm) [Zhao et al 2012]

The first researchers who presented the results about the reduction of friction managed to modify the wall geometry by using the longitudinal grooves called "Riblets". The obtained results indicated a reduction of friction forces of about 8% in comparison with the smooth surface. Riblets of changing geometric shape have been studied on both model scale and full scale to able to observe the effects in resistance reduction. Case ship was modelled in the model sale and ship scale using CFD, and flow calculations were performed without the free surface to simplify the calculations. Patterned surfaces, such as riblets and dimples, have a large potential to be applied into ship surfaces. Two dimensional riblets have been well investigated for aerodynamics and hydrodynamics applications, such as sailing boats. The geometry optimizations are well understood and following conclusions are utilized

• The riblet spacing must be small enough to prevent streamwise vortices in the riblet valleys. s+ =18 is found be the most beneficial separation distance

• Blade type riblets are the most beneficial. A frictional drag reduction up to 10 % can be achieved.

o However the riblet material becomes important as the blade thickness are very small to support the blades

o Riblet height to spacing ratio of 0.5 is found as optimum for blade type riblets

TARGETS

51/94

• Semicircular riblets are the second best achieving up to 6 % frictional drag reduction

o Riblet height to separation ratio should be around 0.7 o Tip thickness and valley becomes important for the production of

the riblets • Triangular riblets can achieve up to 5 % frictional drag reduction

o Riblet height to separation ratio should be around 0.85-0.90 o Triangular riblets are the easiest to produce

• The effects of three dimensional geometry of the riblets are not yet fully understood and techniques for the geometrical arrangement is not developed yet.

• Combination of riblets and hydrophobicity of the riblet blades and valleys must be further investigated. Promising results exists in the literature however further research is needed to reach a decisive results.

• Riblets can also be combined with air lubrication technique. Micro-structured surfaces have beneficial effects on the air lubrication

• The yaw angle for the riblets must be limited to less than 15 degrees. Hence streamlines should be determined for the ship applications and the riblets should be placed according the flow direction.

• Riblet size cannot be directly scaled between the model and ship if Froude scaling is utilised.

• Friction velocity must be calculated for both model and ship scale. Riblet size must be selected accordingly with corresponding non-dimensional height and spacing with average shear stress value. Then resistance reduction ratios can be applied both at model and full scale to obtain frictional resistance with consideration of inflow angle. Wave resistance can be scaled according to Froude method.

• Main problems of riblet application lie on the maintenance of the riblet performance due to fouling and other external effects.

• Under usual environmental and operational conditions use of patterned surfaces cannot be utilised for a robust frictional resistance reduction measure.

TARGETS

52/94

3.6.3 Air lubrication One of the most promising methods to reduce the viscous resistance is air lubrication. Micro-bubbles, air-film and air-cavity techniques have been developed to employ air as a drag reduction medium. Among these three methods, air cavity is the most preferred method to obtain a stabile air layer between the hull and seawater. This report reviews the available techniques for air lubrication, gives details about how air-cavity can be applied into practical ship forms. Applications using CFD techniques are demonstrated. A practical technique to estimate the resistance reduction potential for a conventional ship has also been developed and is presented in the following. CFD techniques were utilised to assess the air lubrication potential on commercial ship hull forms. Five different hull forms, consisting of a tanker, a car carrier and 3 bulk carriers of different size, were utilised to investigate the air lubrication techniques on the commercial ship forms. Both air-film and air-cavity techniques were employed with these forms, with varying air feed quantities. Final assessment of air lubrication among the practical ship forms was made including the energy spent on the air supply. Three physically different techniques are applied to cover the underwater hull surface with air (Figure 43):

• by creating microbubbles in the boundary layer • by creating air film • by creating a recirculation flow fields filled with air

Figure 43: Photo of shark skin surface (scale bar 100 μm) [Zhao et al 2012]

Figure 44: Artificial air cavity ship concept for low speed hull forms

Although a high speed hull bottom can be covered with air by using a single step, this is not possible for a slow ship. Hence a number of cavities are required to obtain sufficient air lubrication area as illustrated in Figure 44.

TARGETS

53/94

Positioning of the steps becomes the most important aspect of such forms. Choi et al (2005) investigates the positioning by CFD. The air cavity chambers can be subdivided not only longitudinally but also horizontally as shown in Figure 45.

Figure 45: Artificial air cavity ship concept for low speed hull forms

The net effect of the air cavity on the viscous resistance is both on frictional and form resistance of the ship. The frictional surface area is reduced due to formation of air recirculation region which in turn leads to frictional drag reduction. Additionally the hull has a virtual extension due to air recirculation region which in turn reduces/increases the form drag of the hull.

TARGETS

54/94

Example case 1: Car Carrier

This type of ship is a high speed vessel which has relatively smaller bottom area with a rise of floor.

Figure 46: Car Carrier Air Cavity Phase Distribution

Example case 2: Bulk Carriers

Bulk carriers are typically slow vessels with an extensive bottom area that offers a good ground on which to apply the air lubrication concept. 3 sizes of bulk carriers have been investigated in this project.

Figure 47: Small Bulk Carrier Air Cavity Phase Distribution

TARGETS

55/94

Figure 48: Medium Bulk Carrier Air Cavity Phase Distribution

Figure 49: Large Bulk Carrier Air Cavity Phase Distribution

Example case 3: Tanker vessel

Same as in the case of bulk carriers apply on tanker vessels. Small speed and large bottom area offer a good proving ground for the air lubrication concept. The fact that frictional resistance is the most important of the resistance components in such vessels renders it’s minimisation even more important.

TARGETS

56/94

Figure 50: Tanker Air Cavity Phase Distribution

By varying the air quantity the following results were obtained for the case studies:

case Speed (kn) Wetted bottom area (m2)

Air Quantity (m3/h)

% gain in viscous resistance

% Power reduction

1 18.5 2271 546 0.0175 1.0 1 18.5 2271 1,093 0.0666 3.8 1 18.5 2271 2,186 0.0875 5.0 1 18.5 2271 3,279 0.1210 7.0 2a 14.5 4000 1,440 0.2599 13.9 2a 14.5 4000 2,880 0.3955 21.1 2a 14.5 4000 5,760 0.4832 25.8 2b 14.5 5343.6 1,440 0.1622 8.7 2b 14.5 5343.6 2,880 0.2600 14.0 2b 14.5 5343.6 5,760 0.3198 17.2 2c 14.5 8310 2,160 0.1788 10.3 2c 14.5 8310 4,320 0.2664 15.4 2c 14.5 8310 8,640 0.3280 19.0 3 14.0 7435 2,880 0.2009 11.3

TARGETS

57/94

3.7 Added resistance

3.7.1 Added resistance due to waves The added resistance due to waves and ship motion were studied in this section. The simulations were conducted for a large bulk carrier and the resulting model serves as an input to the Dynamic Energy Modelling (DEM) that was developed. In order to obtain the added resistance in a certain seaway, a calm water simulation and a simulation with the desired seaway has to be conducted in CFD with the same mesh. The operating condition is the same for both simulations. After calculating the drag coefficients for both simulations the added resistance is obtained as the difference between the ensuing drag coefficients. In this particular case the added resistance is studied for two different drafts. For each draft four different ship velocities are examined with a constant head wave. A series of simulations is conducted for one operating condition with varying regular head waves.

Figure 51: Bulk carrier in three different regular head seas with wave height equal to 3 m (top left: λ=50

m, top right: λ=150 m, bottom right: λ=300 m).

TARGETS

58/94

In figure 51 a bulk carrier is pictured in three different regular head seas. It can be seen that the wave is transported very well to the ship and therefore the calculated added resistance coefficients should be realistic.

Figure 52: Added resistance coefficient ΔcT as function of ship to wave length ratio (LPP/) and

comparison with experiments

As expected the highest added resistance coefficient is obtained for a wave which has a length close to the ship length. The behaviour of the curve is as anticipated for ships with a high block coefficient (CB). The block coefficient of the studied vessel is 0.772. In figure 52 the results are compared to results obtained from experiments (Guo et al. 2011). The experiments were undertaken for a large tanker (LPP=320 m, B= 58 m, CB=0.81), therefore numerical results can only be compared qualitatively to the experimental results. It can be seen that for smaller LPP/λ and close to one, the results are in good agreement. Also for LPP/λ in the interval 3.5 to 5 the results are close to the experimental results. The difference in behaviour for LPP/λ between 1 and 3.5 could be explained by a small difference in the Froude number or the difference of the geometries of the ships. The difference in results for LPP/λ greater than five can either be explained by the reasons above or by the difficulty of calculating the added resistance coefficient for small wave length. It has to be also noted that the experimental data are scattered in the areas in which differences to the results conducted in this study occur. Overall, the added resistance for smaller drafts was found to be higher than the added resistance for higher draft. Froude number has a small influence on the added resistance coefficient. The consistency of the results allows for their use in a resistance calculating model to be included in the DEM software.

TARGETS

59/94

3.7.2 Added resistance of superstructures The superstructures of modern merchant vessels are typically blunt geometries, designed without accounting for aerodynamic performance. The contribution of aerodynamic drag appears small when compared to the total resistance of a vessel. Nevertheless, holistic efforts for increasing the efficiency of shipping need to address and quantify these drag components and the respective optimisation potential as well. This study analyses the aerodynamic characteristics of a tanker that is equipped with a generic deck superstructure. Unlike ship types with a considerable amount of deck cargo as e.g. container vessels, tankers and bulk carriers give more freedom to change the deckhouse dimensions without hindering cargo loading and operation. Following the investigation of the flow characteristics for the base line design, the superstructure of the vessel is retrofitted with different flow guiding devices and their impact on the aerodynamic drag is evaluated. The geometry of the considered flow guiding devices is inspired by publications from the automotive industry, where attempts have been made to reduce the drag of trucks and their blunt geometry. In this study, two different inflow angles are considered, reflecting steaming directly against the wind (1) and at an apparent wind angle of approximately 45 degrees from starboard ahead (2). The atmospheric boundary layer is modelled to account for realistic inflow conditions (Etling 2008). The five different configurations studied are shown in the following figures.

Figure 53: Initial and modified configurations. From top left to low right: Initial setup, setup 1 to 5

TARGETS

60/94

Figure 54 shows a first indication of the computed drag. The base (-suction) pressure is shifted far downstream of the deck house for configurations 2 and 4, which display the largest dead water regime. This has a positive influence on the reduction of the rear-face drag. Moreover rounding the front edges (configuration 5) significantly reduces the extent of the stagnation regime of the front face.

Figure 54: Contour of velocity magnitude shown on slice in xy-plane located at half height of deck house.

Inflow angle 0 degrees. From top left to low right: Initial setup, setup 1 to 5.

In this particular case, the ship’s hydrodynamic resistance is calculated at 1.900E6N. For the baseline design the total aerodynamic drag is 1.728E5 N. Thus the total resistance of the ship is 2.073E6 N with the aerodynamic drag amounting to about 8.3% of the total resistance. For the baseline design, out of the total aerodynamic resistance, approximately 60% relates to the main superstructure, 10% to the bridge wings while the ship’s boottop accounts for the remaining 30%.

TARGETS

61/94

The performance of the different flow guiding devices is assessed by their impact on the aerodynamic drag. Table 7 shows the influence of different flow guiding devices on the longitudinal force (x-force) acting on different parts of the superstructure.

Total Front Rear Wings Baseline 100.0% 100.0% 100.0% 100.0% Setup 1 92.5% 94.2% 84.8% 97.5% Setup 2 86.5% 98.4% 62.2% 102.7% Setup 3 91.0% 97.0% 79.5% 94.3% Setup 4 79.5% 99.9% 41.0% 105.4% Setup 5 61.6% 59.1% 72.4% 41.6%

Table 14: Aerodynamic force (x-component) acting on different parts of the superstructure for different flow guiding devices, 0 degrees inflow angle.

The first three investigated devices reduce the total drag of the superstructure by approximately 10%; the fourth configuration reveals a reduction of 20%. It appears that for these four configurations the x-force of the rear side of the superstructure is drastically reduced, for configuration 4 by 59%; for these configurations the x-force of the front side remains almost constant. The x-force acting on the bridge wings shows fluctuations of about 5% which indicates the expectable accuracy of the results. Configuration 5 gives the highest decrease; here the total deck house drag decreases by almost 40% as does the x-force on the (here rounded) front side. The rounding of the bridge wings reduces the corresponding x-force to 41.6% of their initial value.

TARGETS

62/94

3.8 Renewable sources for propulsion

3.8.1 Sails

Dyna Rig

Already in the 1960ies there was a serious debate on using sail power to propel cargo vessels. Although the emissions problem was not recognised at the time, commercial considerations fuelled the discussion on a possible renaissance of the cargo carrying sailing vessel at a large scale. Hamburg based engineer Wilhelm Prölss (1967) suggested a largely simplified rig that uses an elliptical profile for the mast and fixed profile sails which can be trimmed automatically. The entire rig is self-supporting, with no stays and shrouds necessary.

Figure 55: The “Maltese Falcon” - first large scale Dyna Rig Ship

In a hypothetical case study for a Panamax bulk Carrier of 76000tdw (LPP=217m) under typical operating conditions (v=14kts) the following sail forces were found to be achievable. A sail area of 4400m² at 40m mast height distributed to e.g. 5 or 6 masts is assumed and the resulting forces are shown in figure 56. It is important to note that the model used is somewhat simplified as added resistance due to heel was disregarded. Interaction effects between the masts were not accounted for. No consideration was given to hydrostatic stability, steering ability or structural feasibility. Although the Dyna rig appears to offer superior performance at relatively easy maintenance it still has no major commercial application.

TARGETS

63/94

Figure 56: Achievable thrust over Apparent Wind Angle

Jamda Sails

The JAMDA type wing sail was developed during the 1970s oil crisis by the Japan Machinery Development Association (JAMDA) and Nippon Kokan (N.K.K). It consists of two elements which can be folded about vertical axes for stowing. The entire rig can again be rotated about a vertical axis for trimming. The JAMDA rig was installed on various vessels, the first being the experimental installation on Mini Daigo (83BRT), later on various commercially operated vessel, e.g.:

• Shin Aitoku Maru (600BRT, 200m²), small tanker (series of 17 vessels) • Usuku pioneer (15721BRT, 640m²), bulk carrier • Aqua City (18597BRT), bulk carrier

Two JAMDA sails were installed on all of these vessels and a reduction of fuel oil consumption of 10 to 30% was reported.

Figure 57: Sail configuration of the subject vessel

TARGETS

64/94

In the course of this project a case study was carried out on a contemporary platform (figure 57). The aim is to simulate a merchant ship that is equipped with sails. Both the water and the air phase are considered by employing a volume-of-fluid approach. A tanker vessel is considered as ship geometry since such an arrangement is prohibitive in other types of vessels without major reconstruction. The deck layout considers a deckhouse located in the forward part of the deck which gives an unobstructed sight for the bridge crew. Nine sails are located aft of the deckhouse. Each sail has an area of approx. 1025 square meters. The vessel sails at a velocity of 7.2 m/s and initially at a draft of 17.3 meters at even keel. The sail geometry follows a setup published by Wagner (1966), which is similar to the JAMDA design (Ishihara et al. 1980). The sails have an aspect ratio of 1.66; the sail profile is represented by a circle segment. Over the height the sails do not feature any twist. The geometric particulars are given in the following table.

Chord length l 24.72m Height above deck 10.00m Wing span H 40.00m Aspect ratio A 1.66 Max. camber 2.16m Sail surface (1x) 1025m2

Table 15: Geometric particulars of the sails studied

Figure 58: Flow field around sails visualized by streamlines, traced from 0m, 15m, and 40m above the

lower edge of the sails. Streamlines coloured by magnitude of velocity.

TARGETS

65/94

The formation of tip vortices is shown in figure 58. The use of end plates would reduce the development of tip vortices and thus increase the efficiency of the sail configuration. For the investigated loading condition, the tanker features a large metacentric height. The investigated sail configuration has only a small influence on the floating condition of the vessel. Due to the positioning of the sails along the deck the resulting yawing moment is small, the required rudder angle is below one degree. The associated additional drag is negligible. The comparison of the total drag with the thrust given by the sails reveals a reduction in the required thrust by 11.4%. This reduction in required propulsion power can be compared to the values that are published for retro fitting devices that aim at decreasing the hydrodynamic drag and / or the required propulsion power. Here, usually values below 5% are accepted as worthwhile decrease in resistance and propulsion power. The comparison proofs an encouraging performance of the investigated sail configuration.

The Indo-sail Rig

The IndoSail rig was developed at HSVA by Peter Schenzle during the 1980s within the scope of a development aid project for Indonesia. The vessels developed during this project were intended for short distance coastal traffic in light monsoon wind conditions. Extensive wind tunnel and free sailing tests took place. The selection of technical alternatives for the rigs of a modular series of 3, 4 and 5-masted coastal sailers of between 900 and 2000 t DWT led to the concept of a mechanized roller-reefing gaff-rig with integrated loading gear, which was successfully tested on large free-sailing models.

Figure 59: Indo-Sail scale model tested at the Alster in Hamburg

TARGETS

66/94

The possible reductions in power requirement ensuing from a four-masted variant of the Indo-sail rig are presented in the following. Figure 60 depicts the Achievable sail forces for the considered configuration over apparent wind angles while figure 61 demonstrates the difference of the Indo-sail concept over the Dyna-rig against true wind angles, this time in terms of equivalent power (i.e. scaled to the order of magnitude of a large bulk carrier)

Figure 60: Achievable sail forces over apparent wind angle

Figure 61: Comparison of equivalent power, Dyna-Rig and Indo-Sail

It appears that the Dyna-Rig is advantageous at smaller true wind angles and lower wind speeds, correlating to smaller apparent wind angles. The Indo-Sail rig comes into its own at larger wind angles and higher wind speeds. Generally, the performance envelope of the Dyna-Rig appears more rounded.

TARGETS

67/94

Kite propulsion

The idea behind the deployment of kites for auxiliary propulsion onboard commercial ships is quite straightforward: a kite has a single-point connection to a derrick at the forecastle area of a ship and through a tow line it transmits the wind loads into an additional propulsion force (horizontal force component) to the ship (figure 62). The kite is fabricated by high strength textile and it is fitted with air cavities in order to maintain its shape and its performance.

Figure 62: Kite application in a small cargo ship

(www.skysails.com) Figure 63: Relative fuel saving for tanker of the

case study

In particular, the system is comprised by (i) a towing kite with the rope, (ii) a launch and recovery system, and (iii) an automatic control system. Except for the main parts, a multitude of subsystems is also present to ensure the smooth, automated and reliable operation of the kite. This method of capturing the wind energy has a certain range of advantages compared to the more conventional sails, which are deployed on masts and on the ship weather deck.

- The kite can be actively controlled in order to create its own flying speed and therefore affect the direction of the apparent wind, and in turn, increase the produced additional thrust;

- The fact that the kite is operating in relatively high altitude above the water surface implies that it is working at the higher levels of the wind boundary layer in the open sea and therefore it is exposed to stronger winds, i.e. it is more efficient;

- Because of the single-point connection to the ship, the induced motions to the ship are minimal;

TARGETS

68/94

- No masts are needed for the kite deployment and therefore no deck space is used.

Despite these advantages, the kite operation requires a certain range of control, a feature that is integral to the overall system, and customisation to a particular ship. The kite design and configuration is conducted for a range of wind speeds and directions as it will be explained in the following sections. That is, the efficiency of a kite system is directly linked not only to the prevailing wind conditions but also on the rapidity in which they change. This inherent characteristic of the system makes many ship operators reluctant to install it. Application of this theory in the case of a 50,000 DWT tanker, with a kite of 500 m2 area, and towing line 350 m long, shows fuel savings up 50% (figure 63). It should be stressed that the results of this study require special attention as the theory has not been validated by sea trials or any other method up to the day of the preparation of this report. Similar statements can be found by other authors in the field. Most importantly though, the complexity of the kite deployment and operation has not received adequate attention by the academic community so far. As a result, the approaches to model its performance at theoretical level, and therefore prove its added value, are limited. The main reason for this situation is that full scale operational data are not available at the moment hence the existing models are still under development. Nevertheless, the kite system has been installed onboard four ships and a test plan is well underway.

TARGETS

69/94

3.8.2 Flettner rotors The physical effect of fast rotating bodies generating lift was discovered by Newton in 1672 and correctly described by Magnus in 1852. When a body (such as a sphere or circular cylinder) is spinning in a viscous fluid, it creates a boundary layer around itself, and the boundary layer induces a more widespread circular motion of the fluid. If the body is moving through the fluid with a velocity V, the velocity of the thin layer of fluid close to the body is a little greater than V on the forward-moving side and a little less than V on the backward-moving side. This is because the induced velocity due to the boundary layer surrounding the spinning body is added to V on the forward-moving side, and subtracted from V on the backward-moving side.

Figure 64: "Magnus” effect on a rotating cylinder. Figure 65: ENERCON E-Ship 1, 2010.

In 2007 the German wind turbine manufacturer ENERCON started to develop a new ship for waterborne transport of their wind turbines. The ‘E-Ship 1’ evolved into a showcase design for energy efficiency technologies, sampling the most promising propulsion technologies combined with a fully optimised hullform developed from numerical and experimental analyses. As a prominent feature, the E-Ship 1 sports 4 Flettner Rotors mounted behind the forward-placed superstructure and at the stern of the vessel as indicated in the illustration below. The vessel is in operation since 2010.

Currently two companies are building or intending to build Flettner rotors as auxiliary propulsion for vessels: Greenwave (www.greenwave.org.uk) and WindAgain (www.windagain.com). Interestingly, the claims on fuels saving resp. generated power are widely different. While Greenwave claim fuel savings of about 13%, respectively 1000t HFO or 3000t CO2 per year on a small bulk carrier, WindAgain claim savings of 20 to 25% for a Panamax bulker.

TARGETS

70/94

Figure 66: Potential application of Flettner rotor on a bulk carrier.

The following figure depicts the achievable wind-derived driving force respectively power saving for a Panamax bulker (76000tdw, LPP=217m), similar to the one shown above, over a range of true wind angles and speeds. The vessel’s speed is kept constant at 14kts, corresponding to a hull drag of 640kN (from tank test data). Rotation rates are optimised for minimum total power requirement (main propulsion + rotors).

Figure 67: Achievable thrust from three rotors over Apparent Wind Angle.

TARGETS

71/94

3.9 Auxiliary power generation The use of renewable energy sources for auxiliary power generation is more essential than ever. With emissions’ regulations ever tightening, ships must reduce their fossil fuel consumption to as low as practically possible. Of course at the moment such alternative technologies are at their beginning, meaning that there are inefficient and difficult to install and operate compared to traditional installations. However, the more they are being applied the more likely it is that the time will come that they will serve as a feasible alternative. In addition, operators that choose to invest in such technologies at the moment benefit from advantageous regulations and have the added profit that they are in fact ahead of their game. That said, there are also significant gains to be harvested from alternative applications, with most technologies having a short enough return of investment period, especially with the rising prices of bunker cost. A few of the available technologies have been studied in this project and will be presented in the sections to follow.

3.9.1 Photovoltaic installations Photovoltaics describe the conversion of solar radiation into electrical energy. The so-called photoelectric effect was already discovered in 1839 and since then has formed the basis of the photovoltaic Energy-production. A solar cell or photovoltaic cell is an electrical component, the short-wave radiant energy, usually sunlight, directly converts into electrical energy. The physical basis of the conversion is the photovoltaic effect, which is a special case of the internal photoelectric effect. Under laboratory conditions degrees of efficiency of up to 40% have been reported, but are currently not realistic for industrial applications. For maritime applications some advantages in comparison to land-based installations have been observed. It has been proven that degrees of efficiencies have been higher when using PV-installations at sea. To the present day, several projects using photovoltaic systems have been realized. Most of these built vessels are small boats, sailing on rivers or lakes.

Figure 68: Functional principle of a solar cell

TARGETS

72/94

However, there are also a few sea-going vessels, using solar radiation as an energy source, both cargo and passenger vessels. Since the electrical board network uses alternating current, the photovoltaic cells have to be combined with an alternating-current converter in order to be connected to the AC-board-network. Most installations on current ships contribute to less than 1% of the total energy requirement and are mostly limited by space where the system can be installed. However, new designs that are specifically developed for the use of PV systems offer contributions of up to 3% of the total energy needs. A case study was carried out in the course of this project, utilizing a car carrier as the test module. Figure 69 following, shows the potential position of the PV units on the vessel. The evaluation of the constructible surface, according to the GA-Plan, resulted in an area of 2170 m².

Figure 69: Potential Areas for the installation of PV-Units.

The system is suggested to be horizontally mounted, which is surely not the solution (which would be an alignment directly facing south) but represents the best compromise without any tracking device. Tracking systems would lead to a higher demand of installation spaces which would not be an economical solution. The predicted peak power of the whole PV-System is about 236.5 kW, based on reference value of 0,109kWP/m². An investigation on the annual energy demands of the subject vessel based on route information as well as operational parameters such as speed, loading and time at sea resulted in a calculated annual demand of approximately 7225MW. The simulations resulted in an average sharing of PV-Load of 5% and an environmental impact as shown in table 16.

TARGETS

73/94

Average sharing of PV load [%] 5 Annual reduction potential of fuel [tn] 203.1 Annual reduction potential of Carbon dioxide [tn] 649.63 Annual reduction potential of Carbon monoxide [tn] 1.47 Annual reduction potential of unburned hydrocarbon [tn] 0.15 Annual reduction potential of particulate matter [tn] 0.1 Annual reduction potential of Sulphur oxides [tn] 18.54 Annual reduction potential of Nitrogen oxides [tn] 12.34

Table 16: Environmental impact of the PV installation

3.9.2 Fuel cells

General information

Fuel cells are electrochemical devices that convert chemical energy into electricity. Originally hydrogen and oxygen were used for the controlled redox reaction, which generates a cell voltage, comparable to the voltage of a voltage source. In order to sum up the single cell voltages, the cells need to be connected serially. This is commonly known as stacking and leads to the cell stack, which is a series of connected cells. The fuel cell process is not subject to the efficiency limitation of the Carnot factor. Power alternation in a fuel cell is not done via thermal energy transmission. The Carnot factor describes the highest possible value of thermal efficiency for a thermodynamic cycle with respect to the ambient temperature. The following figure describes the differences between an internal combustion engine (ICE) and a fuel cell process.

TARGETS

74/94

Figure 70: Power alternation for internal combustion engines (above) and for fuel cells (below).

Because of the fact that ICE’s performance is proportional to working temperature the fuel cell efficiency can exceed the Carnot limit at low working temperatures. This is because the electrochemical process does not involve a conversion of thermal to mechanical energy. The difference is shown in figure 71.

Figure 71: Theoretical comparison between the Carnot Factor ad Fuel Cell Efficiency

TARGETS

75/94

In principle, fuel cells are created equally. Between two electrodes, the anode and the cathode, an electrolyte is placed. The electrolyte allows hydrogen ions to travel from the anode side to the cathode. The electrodes are connected via an external circuit. When the electrical circuit is closed, the electrons flow via a consumer to the cathode. The basic structure of a fuel cell can be reduced to cathode, anode and the electrolyte. The above mentioned reactions take place at the corresponding electrode. In short there are six types of fuel cells available, each with their advantages and disadvantages. These are:

• Alkaline (AFC) • Proton exchange membrane (PEMFC) • Direct methanol (DMFC) • Phosphoric acid (PAFC) • Molten Carbonate (MCFC) • Solid oxide (SOFC)

In general, every hydrocarbon-based fuel can be used as a fuel cell fuel, but all of these fuels have to be reformed into hydrogen to work in an electrochemical cell. Using pure hydrogen is the desired way, but nowadays there is no existing infrastructure for hydrogen supply. For naval application pure hydrogen, natural gas and methanol appear to be suitable. With a hydrogen tank a simpler FCS can be installed onboard, as no complex fuel treatment is necessary for operation. A comparison of costs for hydrogen and MGO is given. In May 2011, 1 kg of hydrogen cost €8.099 at a hydrogen petrol station in Berlin. The energy-specific costs for hydrogen are:

𝑐𝑜𝑠𝑡𝑠𝑒𝑛𝑒𝑟𝑔𝑦

=€8.099120𝐺𝐽

= 0.067€𝐺𝐽

In March 2013, MGO reached a price level of 913$/ton (Petromedia LTD, 2013). Therefore the energy-specific costs are:

𝑐𝑜𝑠𝑡𝑠𝑒𝑛𝑒𝑟𝑔𝑦

=€0.7042𝐺𝐽

= 0.017€𝐺𝐽

The comparison shows that hydrogen is four times more expensive than MGO. A lower hydrogen price can be expected, when bunkering higher hydrogen amounts.

TARGETS

76/94

Various applications

Fuel cells were introduced to the public during their use in US space technology in the 1960s. For example, during the Apollo programme one spacecraft was equipped with three alkaline fuel cells, in order to provide electrical power for operation and for water supply. Naval application started in the 1960s. The first ship equipped with a fuel cell is said to be the submersible “STAR I” (1964), a small one-man submarine, owned by the US defence group General Dynamics. Like in the above-mentioned Apollo spacecraft an alkaline fuel cell was used to provide electrical power. In the last 50 years fuel cells have been used for many pilot projects in the maritime sector as propulsion systems or on-board power supply. However, except for the German Submarine “Class 212A”, which is equipped with a PEMFC for propulsion, we still talk about pilot projects for naval use. The fact though that a submarine is equipped with such a system means that, if designed properly, any issues regarding space and operation should be overcome. In the maritime sector some convincing projects have taken place. Since 2008 a hydrogen-operated passenger vessel has been cruising on Lake Alster and the surrounding channels. The vessel (figure 72) is 25.6m long, 5.2m wide and has a maximal draught of 1.2m. The ship is equipped with two fuel cells, 48kW each, and a lead gel battery pack. In one of these FCSs six stacks of 96 cells each are put together. The combination allows the ship to speed up to 15km/hr. In an onboard tank 50kg of pressurized hydrogen (350bar) are stored, providing fuel sufficient for three days. "The Naval Architect" published a comparison between a conventional and the fuel cell driven inland vessel in 2008. An equivalent inland ship would reach emissions of 1000kg NOx, 220kg of SOx, 40kg of particulates and 72500kg of CO2 per year. The fuel cell driven "Alsterwasser" emits during operation only H2O and therefore all these values fall to zero.

TARGETS

77/94

Figure 72: Construction of the "Alsterwasser".

Another experimental application is the Viking Lady. An offshore supply vessel, she was designed to carry an MCFC fuel cell from Germany´s MTU Onsite Energy. The fuel cell is located in a container and in combination with an inverter the power is used for onboard energy supply. The system is designed in such a way that the 320kW fuel cell power meets the needs of the necessary shipboard service power demand. During standby and harbour operations the power delivered from the fuel cell is high enough to supply all electrical consumers. The MCFC fuel cell works with a temperature of 600°C. The exhaust gases leave the system with a temperature of approximately 400°C. This waste heat (250kW) of the fuel cell plant is used as well. A heat exchanger extracts the thermal energy from the exhaust gas and transfers it to a heating system, in which also the waste heat of the main engines is captured.

Figure 73: The FCS is placed on the main deck in a container. In service the system is stored inside the

accommodation opening.

TARGETS

78/94

Concept with hydrogen as fuel

While hydrogen is used as a fuel cell fuel there is no need for a reformation process. Hydrogen is stored in a tank and fed into the FCS via a control valve. The hydrogen reacts with oxygen to the water inside the cell. According to the power demand, a fan has to provide a sufficient air supply for this reaction. The direct current is fed into the grid by an DC-AC inverter.

The fuel supply model (1) includes the following systems: • a pressurised hydrogen tank • if necessary a hydrogen compressor

The fuel cell system (2) contains: • if suitable, a high temperature cooling water supply for preheating the

stack and cooling the fuel cell during operation • a hydrogen supply including a control valve for operation • a fan for oxygen supply during operation • a control device which ensures proper cooperation of all subsystems

The grid connection (3) consists of: • a DC-AC inverter • a power switch

Concept with methanol as fuel

As the methanol fuel needs to be reformed to hydrogen this solution is more complex. The methanol will be stored in a bunker tank and has to be mixed with demineralised water in a determined ratio. Therfore, a control device ensures the necessary volume flow of each liquid. Also the fuel demand for the FCS, which is in some way proportional to the electrical load, has to be delivered from this system. Otherwise there can be a solution where the methanol hydrogen mixture is stored in a mixing pipe, which is located above the FCS in order to provide a sufficient supply pressure. The FCS is connected with an internal steam supply in order to run the reformation process, where hydrogen and carbon dioxide is transacted from the mixture. The hydrogen is fed into the fuel cell in order to generate a current. The direct current is fed into the grid via a DC-AC inverter.

The fuel supply model (1) contains several systems: • a mixing tank located above the fuel cell

TARGETS

79/94

The fuel cell system (2) includes the following interfaces: • if suitable, a connection / interaction to the steam system in order to

provide energy for the reformation process • a connection to the high temperature cooling water to get energy for

heating up the fuel cell stack and for cooling during operation • a fan in order to provide sufficient oxygen supply for the FCS in

accordance with the deliverd electrical power • a control device which arranges an efficient actuator motion

The grid connection (3) is made up of the following subsystems: • a DC-AC Inverter • a power switch

3.10 Energy storage for marine applications

3.10.1 Necessity of energy storage Energy storage is essential, especially for the integration of renewable and conventional energy sources. Such is the nature of renewable sources that the availability is not perpetual. Thus, electricity storage can enhance the value of energy from renewable generation in at least two fundamental ways. Storage can “firm-up” renewables’ output so that electric power (kW) can be used when needed. Similarly electric energy (kWh) generated when the relevant source is available can be “time-shifted”. Storage can also save energy by reducing peaks in demand by “shaving” peak loads.

3.10.2 Types of energy storage devices The so-called “Ragone” chart represents the energy storage technology according to the energy-to-weight ratio and the power-to-weight ratio (Figure 74) Thanks to this chart; it is easier to choose the energy storage needed.

TARGETS

80/94

Figure 74: Ragone overview

Conceptually, the vertical axis describes how much energy is available, while the horizontal axis shows how quickly that energy can be delivered, otherwise known as power, per unit mass. For example, powering a small light-bulb may require low amounts of power, but the power should be delivered slowly enough to operate a flashlight for minutes or hours of use. Conversely, a high speed electronic switch inside a computer may require very little energy to activate; yet it must be delivered rapidly enough to complete the transaction in mere microseconds. These two types of loads would be represented at opposite corners of the Ragone chart. The choice of ESD is primarily based on cost of application, return period and its position in the Ragone Chart as shown in figure 74. The following types of energy storage can be identified:

Chemical storage:

This type comprises of three sub-types, namely the primary batteries, which are charged during the manufacturing process and cannot be loaded again in general; accumulators, which can be set, by a corresponding amount of energy, back to their original condition and tertiary batteries, such as fuel cells, that they are continuously supplied by a chemical process and respective chemical products and electrical energy are taken out as a product.

Mechanical storage:

One of the most important sub-types of mechanical storage, with various industrial applications is the flywheel. Used for purposes like evening out discontinuous energy sources or controlling the orientation of spinning objects.

TARGETS

81/94

They store kinetic energy in a high speed rotating drum which forms the rotor of a motor generator. When surplus electrical energy is available it is used to speed up the drum. When the energy is needed the drum provides it by driving the generator. Modern high energy flywheels use composite rotors made with carbon-fibre materials. The rotors have a very high strength-to-density ratio, and rotate at speeds up to 100,000 rpm in a vacuum chamber to minimize aerodynamic losses. The use of superconducting electromagnetic bearings can virtually eliminate energy losses through friction. Energy stored in flywheels spans from 2.7kWh in a small flywheel battery to 26kWh in electrical power backup flywheels. The efficiency of flywheels can be as high a 90%. The second sub-type of mechanical storage to describe is the compressed air storage. The physical law behind this idea is the “incompressibility of gases” which states, that it is not possible to shrink the volume of a gas by applying force on it. But the law of “conservation of energy” has to be fulfilled too, therefore the gas is pressurized. This process generates a lot of heat, while on the opposite, decompression requires heat. Storing the heat emitted during compression and using it again during decompression, also considerably increases the efficiency of the process. The basic idea behind it is to support the electricity supply chain. At times of high electricity availability air is compressed and the resulting heat is placed in an interim heat-storage device. When electricity demand rises, this compressed air can be used to generate power in a turbine – while recovering the heat. Typical systems can reach a power density of 1kW/kg with an efficiency of 65%. Another type of mechanical storage is the electro-magnetic. Superconducting magnetic energy storage systems (SMES) store energy in the field of a large magnetic coil with direct current flowing. It can be converted back to AC electric current as needed. Low temperature SMES cooled by liquid helium is close to be commercially available. High temperature SMES cooled by liquid nitrogen is still in the development stage and may become a viable commercial energy storage source in the future. SMES systems are large and generally used for short durations, such as utility switching events. Power density is similar to compressed air at 1kW/kg but efficiency is an almost perfect 99%. They are extremely expensive and have, therefore, no commercial application yet.

TARGETS

82/94

3.11 Integration of energy modules

3.11.1 Dynamic Energy Modelling Probably the most important thing in any attempt to reduce anything is to be able to quantify it. Energy is no different in this respect, thus, identifying a big gap in the measurement of energy consumption in vessels, an attempt was made in this project to create an integrated tool for this purpose. The tool integrates a number of modules that were developed each for the purpose of quantifying the performance of a specific part of the vessel’s consumers. In this respect, modules were developed for all the major consumers of the ship such as resistance, propulsion, electrical and mechanical installations and various alternative means of energy efficiency. The advancement from state-of-the-art is that for the first time all the vessel’s major consumers are modelled and integrated alongside things that deteriorate the vessel’s performance over the years, such as fouling of the hull and propeller. This will be made in a dynamical manner so that the vessels performance can be simulated in various conditions (loading, weather etc.) over a large period of time so that ways to improve the energy efficiency can be identified. The modelling of dynamical systems can be made either sequentially (modular) and simultaneously. In the sequential approach the output of every module is a function of (a) the input it receives, and (b) its internal parameters. That is, the output of one component becomes the input to the next component and vice versa. This approach is advantageous when the components of a specific energy system can be replaced from a library of various predefined components, (Colona, et al., 2007). The interactions between components (i.e. input/output ports) can be predefined (causal scheme) or not (non-causal scheme). In the present study the sequential (modular) approach with causal interactions will be used. In some cases the non-causal interaction between components will be unavoidable. In the simultaneous approach all the equations that define a system and its various components are treated as a whole. This is a much faster and more computationally efficient approach because the equation set can be simplified and optimised for the particular solver/software that is used. The simultaneous approach will also be introduced. In essence the modular approach can be used in concept exploration studies (e.g. preliminary design), where the design space of feasible and yet attainable solutions can be properly explored with parametric studies. The simultaneous approach is better suited for the performance appraisal of existing energy system configurations, pertaining to the operational and retrofitting stages of

TARGETS

83/94

the ship under consideration. With the simultaneous approach, a range of studies can be conducted according to a set of performance metrics (fuel consumption or amount of exhaust gas emissions) for the identification of poorly performing parts/sub-systems, control schemes suitability, etc. It is the interaction of the various sub-systems mentioned earlier that makes it so hard to quantify efficiency. All individual methods of energy saving, promise a reduction of consumption by a certain amount. Practise, however, has shown that these amounts don’t act collectively, much to the confusion of the industry. It is only through an integrated approach that the overall reduction of consumption can be estimated.

3.11.2 Energy systems modelling The energy domains that have been considered in TARGETS are thermal, electrical and mechanical. In some cases, depending on the context, thermal energy has been subdivided into sensible and latent energy and the mechanical energy has been subdivided into kinetic, potential and pressure flow energy. By definition energy can only be stored, transferred, and converted. Thus, energy systems and components can only store, transfer and convert energy between various types. In the current context, a system is defined as a combination of components, which performs a task indicated by a function. A system can be decomposed in a hierarchical manner into subsystems and components (including their physical connections). The focus of the analysis relies on the modelling and verification of the individual subsystems and components. In the modelling process of a system the following need to be identified:

• Whether each individual system/component performs energy conversion, transmission and / or storage;

• The system boundaries; • The input, output and the states which express the behaviour of the

system/component, and • The specific physical characteristics and constraints of the system (rated

power, number of operational states, weight, maximum rpm etc.).

The boundaries of the system can be either physical or abstract. The input to a system (or its constituents) can be subdivided into (i) control (input that can be controlled, e.g. control systems), (ii) disturbances (input that cannot be controlled e.g. environment) and operational (input influenced by ship’s operations e.g. cargo carried). At component level, the input and output should

TARGETS

84/94

be either potential variables (temperature, voltage, pressure, torque, etc.) or flow variables (volume flowrate, mass flowrate, electric current, etc.). In some cases the input and output may involve non-physical variables like enthalpy, entropy, information flow, etc. A component can be further decomposed into many levels depending on the underlying phenomena that describe its behaviour. From a simulation point of view, a component is regarded as the indivisible building block of an energy system and it may be physical in nature or fictitious. Its internal behaviour, i.e. the state of the component depends on input and/or output and can be modelled as a set of differential and/or algebraic equations, an explicit function, tabular data, etc. By means of an example, the interaction of the main components associated with the propulsion system (i.e. main engine, shaft, propeller and the ship itself) is shown in figure 75.

Figure 75: Simplified ship propulsion system

Ship resistance

Ship resistance is hence the prime factor determining the power requirement for the vessel’s main engine. The exact prediction of the resistance is a key input to an accurate determination of the vessel’s power requirement. Detailed analyses decompose the ship’s total resistance into a number of different components, each describing a significant source. The viscous resistance forces are generated through surface friction on the ship’s hull. The pressure forces, often called residual resistance, are form related, including dynamic pressure gradients and wave generation. During service, a number of additional operational components such as wind, waves and current need to be addressed in the resistance prediction analysis. A simplified decomposition of ship’s resistance is shown in figure 76.

TARGETS

85/94

Figure 76: Simplified ship resistance decomposition

These components can be predicted for the design condition in calm water as well as for off-design conditions with a significant environmental influence from wind and waves. RANS computations provide a more precise and accurate prediction of ship resistance. However, these RANS computations are still very time-consuming, because each variation in operating conditions requires a new computation and in most cases a new computation mesh. Traditional estimations have been the preferred method, as they are significantly faster for the level of accuracy they deliver. The choice of prediction method is strongly influenced by the amount of ship-specific detail and its operating condition for evaluation, as well as the time available for the level of accuracy needed.

TARGETS

86/94

Individual systems

Furthermore, a number of systems have been developed for the basic functions of the vessel, such as the fuel oil system (figure 77), the function of which is to provide fuel oil to internal combustion engines and auxiliary boilers and the lube oil system, the function of which is to provide lube oil to internal combustion engines.

Figure 77: FO System Decomposition

Other systems are the cooling water system, the sea water / central cooling water system, the compressed air system, the ventilation system, the refrigeration system, the fresh water generation and distribution system, the electric power system, the steam and feed water system and the ballast water system. Moving to the accommodation areas the biggest consumer is the heat, ventilation and air condition (HVAC) system (figure 78). The latter is a complex system which includes specific machinery, ducts, pipes and fans. Its purpose is to enhance the quality of the air and circulate it in certain spaces of the ship in order to provide living and reasonable working conditions for the crew on board. The two factors that determine air quality are temperature and relative humidity which subsequently are the two main controlled inputs of the HVAC system. However, the continuously changing environmental disturbances (e.g. the ambient temperature, air humidity, wind speed and direction, solar radiation) directly affect the temperature and relative humidity of the air conditioned spaces and therefore need to be taken into account. Also apart from the environmental disturbances, data for the occupancy profile and the

TARGETS

87/94

existence of any heat generating devices in the targeted spaces (internal disturbances) need to be taken into consideration since they affect the output of the system.

Figure 78: HVAC System Decomposition

Then there are the mission specific systems, which relate and are unique to a specific type of vessel. For example for tanker vessels there are the tank heating and cleaning systems, the inert gas and ventilation system and the pump room ventilation system. For container vessels there are the reefer electric power sub-system, the function of which is to provide electric power to refrigerated (reefer) containers and the cargo hold ventilation system. For Ro-Ro and Ro-Pax vessels there is the car deck ventilation system and the deck hydraulic system for the operation of ramps. Finally for bulk carriers there are the deck hydraulic system for hatch operation, the hydraulic system for deck crane operation and the cargo hold cleaning system. All these systems have been modelled and contribute to the vessels total consumption through the integration platform.

TARGETS

88/94

4 Conclusions Energy efficiency of ships is an extremely complex problem. Various attempt made in the past had a common drawback: integration or rather, lack of it. In this project, an integrated approach was undertaken for the first time with spectacular results. Not only new developments were made for alternative means of energy saving but they were also modelled, quantified and integrated into a broader platform that can estimate the consumption of the vessel on a dynamic basis. The deterioration of a vessel’s performance was modelled for the first time in order to be able to plan a more efficient maintenance schedule and also added to the integration platform. The savings achieved by each energy-saving device were measured individually but also collectively in order for the designer/engineer to be able to choose the most appropriate solutions for each discrete application. It was also established that interactions between the various energy-saving devises compromise their performance, necessitating the careful planning of their application.

TARGETS

89/94

5 Bibliography and References Candries, M (2001) Drag, Boundary Layer and Roughness Characteristics of Marine Surfaces Coated with Antifoulings, Ph.D. Thesis, University of Newcastle Upon Tyne, Newcastle, UK.

Colona P. and van Putten H., Dynamic modelling of steam power cycles. Part I - Modelling paradigm and validation, Applied Thermal Engineering, 2007, Vol. 27.

Etling, D.: Theoretische Meteorologie: Eine Einführung. Springer Verlag, 3rd Edition 2008.

Guo B., Steen S.: Evaluation of added resistance of KVLCC2 in short waves. Journal of Hydrodynamics, 2011.

Ishihara M., Watanabe T., Shimizu K., Yoshimi K. and Namura H.: Prospect of sail equipped motor ship as assessed from experimental ship “Daioh”. SNAME Shipboard Energy Conservation Symposium 1980.

Schneekluth H., Bertram V.: Ship Design for Efficiency and Economy. Butterworth-Heinemann, 1998

Wagner, B.: Windkanalversuche mit gewölbten Plattensegeln, mit Einzelmasten sowie mit Plattensegeln bei Mehrmastanordnung. Schriftenreihe Schiffbau, Bericht 171, Institut für Schiffbau der Universität Hamburg. Hamburg 1966.

Wilhelm Prölss: Windkanalversuche für einen sechsmastigen Segler nach Prölss, Schiff und Hafen, 3/1967 (in German)

Willsher, J. (2007), “The Effect of Biocide Free Foul Release Systems on Vessel Performance”, International Paint Ltd., London/UK, 8 October 2007

Zhao D-Y, Huang Z-P, Wang M-J, Wang T., Jin Y., 2012, Vacum casting replication on shark skin for drag-reducing applications, Journal of Materials Processing Technology, 212, 198-202

TARGETS

90/94

6 Indexes

6.1 Index of Tables Table 1 List of ESPs suitable for construction ................................................... 7

Table 2 List of ESPs suitable for retrofitting .................................................... 13

Table 3 Fouling degree (Source: Schulze) ....................................................... 17

Table 4: Obtainable load range after one T/C is cut-out ................................. 18

Table 5 List of ESPs for operation ................................................................... 20

Table 6 Guidance on the Number of D/Gs in Operation .................................. 23

Table 7: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the adaptation of the bulbous bow of the 8,500 TEU container vessel to slow-steaming conditions .............................................................................. 43

Table 8: Daily savings of energy, fuel oil, CO2 emissions and operational costs through pre-trimming the 8,500 TEU container vessel (3m bow down) ........... 43

Table 9: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the modification of CAP SAN NICOLAS’ bulbous bow and accumulated savings for the TARGETS voyage data ............................................................. 44

Table 10: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the aftship modification for a bulk carrier and accumulated savings for the TARGETS voyage data .............................................................................. 45

Table 11: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the fitting of a Schneekluth nozzle for a small bulk carrier ................ 45

Table 12: Daily savings of energy, fuel oil, CO2 emissions and operational costs through the adaptation of the bulbous bow of the 8,500 TEU container vessel to slow-steaming conditions .............................................................................. 47

Table 13: Effect on skin friction as compared with the hydrodynamically smooth hull ............................................................................................................... 47

Table 14: Aerodynamic force (x-component) acting on different parts of the superstructure for different flow guiding devices, 0 degrees inflow angle. ...... 61

Table 15: Geometric particulars of the sails studied ....................................... 64

Table 16: Environmental impact of the PV installation .................................... 73

TARGETS

91/94

6.2 Index of Figures Figure 1: Use of propulsion energy (IMO, 2009) ................................................ 5

Figure 2: Contra-rotating propellers ................................................................ 8

Figure 3: Axe-bow principle for reduction of wave-induced resistance ............. 9

Figure 4: Effect of Part-load Optimization in Specific Fuel Oil Consumption of Main Engine (Source: MAN Diesel) ................................................................ 10

Figure 5: Wake distribution of the asymmetric stern ....................................... 12

Figure 6: Propeller Forces Acting on Rudder (left), Installed Combination of Costa Bulb and Transversal Fins on Tanker (centre) and Costa bulb (right). ..... 14

Figure 7: Fluid Flow at Stern where a Mewis Duct is Present ............................ 15

Figure 8: Detailed Schematic of Mewis Duct System (Source: Becker Marine Systems) ........................................................................................................ 15

Figure 9: Underwater propeller polishing........................................................ 17

Figure 10: Turbocharger cut out with a swing gate valve (Source: MAN B&W) .. 18

Figure 11: Cylinder cut-out system ................................................................ 19

Figure 12: Investment return period (in months) ............................................ 21

Figure 13: Typical Vessel Operational Profile by Mode .................................... 22

Figure 14: Results without propulsion. Contourplot of sensitivities (left) at a Reynolds number of 1E7 (red [blue] indicates a reduction [an increase] of displacement). Corresponding wake field upstream of the propeller disk (right). ........................................................................................................... 25

Figure 15: Results with propulsion. Contourplot of sensitivities (left) at a Reynolds number of 1E7 (red [blue] indicates a reduction [an increase] of displacement). Corresponding wake field upstream of the propeller disk (right). ........................................................................................................... 25

Figure 16: Free surface elevation calculated with original scheme and parameters. ................................................................................................... 26

Figure 17: Free surface elevation calculated with original scheme and parameters. ................................................................................................... 26

Figure 18: Bulb modification. The blue lines show the original hull. ................ 28

Figure 19: Total resistance - comparison of different bulb geometries at T = 13.m ............................................................................................................. 28

Figure 20: Hull Pressure and wave elevation at the bow, T = 11.5m, v = 12 kts ..................................................................................................... 29

Figure 21: Comparison of wave pattern - Catamaran (left) and Monohull (right) ..................................................................................................................... 29

Figure 22: Resistance coefficients for a catamaran, monohull and trimaran with equivalent displacement ................................................................................ 30

TARGETS

92/94

Figure 23: Total Resistance - Comparison of Trimaran, Monohull and Catamaran of equivalent displacement ........................................................... 31

Figure 24: Adjoint Sensitivities, Profile View (upper) and aft view (lower) ......... 32

Figure 25: Optimisation Results for Propeller Performance ............................. 32

Figure 26: Axial Component of Propeller Wake, Baseline (left), Improved Skeg (right) ............................................................................................................ 32

Figure 27: Original aft hullform ...................................................................... 33

Figure 28: Modified aft hullforms ................................................................... 33

Figure 29: Effect of wedge height on resistance.............................................. 34

Figure 30: Aft lines plan of studied vessel with indication of duct and propeller ....................................................................................................... 34

Figure 31: Tip-Raked propeller blade ............................................................. 36

Figure 32: Comparative gain in efficiencies for a downstream (post) stator/propeller and upstream (pre) stator/propeller arrangement including the effect of stator blade number......................................................................... 37

Figure 33: Post-swirl stator arrangement for large bulk carrier ....................... 37

Figure 34: The BLAD concept. Close-up view of the starboard deflector and position of the deflectors. .............................................................................. 39

Figure 35: Container Ship: Location of Appendages ........................................ 40

Figure 36: Effects of Trim on powering, Mean Draft 11.5 m ............................ 41

Figure 37: Effects of Trim on powering, Mean Draft 13.0 m ............................ 42

Figure 38: Main types of fouling organisms .................................................... 46

Figure 39: Computational domain and grid setup for the test vessel ............... 48

Figure 40: Normalised resistance components as function of surface roughness (computed and ITTC) ..................................................................................... 49

Figure 41: Friction coefficient on smooth (upper) and rough (lower) hull ......... 49

Figure 42: Photo of shark skin surface (scale bar 100 μm) [Zhao et al 2012] ... 50

Figure 43: Photo of shark skin surface (scale bar 100 μm) [Zhao et al 2012] ... 52

Figure 44: Artificial air cavity ship concept for low speed hull forms ............... 52

Figure 45: Artificial air cavity ship concept for low speed hull forms ............... 53

Figure 46: Car Carrier Air Cavity Phase Distribution ........................................ 54

Figure 47: Small Bulk Carrier Air Cavity Phase Distribution ............................. 54

Figure 48: Medium Bulk Carrier Air Cavity Phase Distribution ......................... 55

Figure 49: Large Bulk Carrier Air Cavity Phase Distribution ............................. 55

Figure 50: Tanker Air Cavity Phase Distribution .............................................. 56

Figure 51: Bulk carrier in three different regular head seas with wave height equal to 3 m (top left: λ=50 m, top right: λ=150 m, bottom right: λ=300 m). 57

TARGETS

93/94

Figure 52: Added resistance coefficient ΔcT as function of ship to wave length ratio (LPP/) and comparison with experiments .............................................. 58

Figure 53: Initial and modified configurations. From top left to low right: Initial setup, setup 1 to 5 ............................................................................... 59

Figure 54: Contour of velocity magnitude shown on slice in xy-plane located at half height of deck house. Inflow angle 0 degrees. From top left to low right: Initial setup, setup 1 to 5. ..................................................................... 60

Figure 55: The “Maltese Falcon” - first large scale Dyna Rig Ship .................... 62

Figure 56: Achievable thrust over Apparent Wind Angle .................................. 63

Figure 57: Sail configuration of the subject vessel .......................................... 63

Figure 58: Flow field around sails visualized by streamlines, traced from 0m, 15m, and 40m above the lower edge of the sails. Streamlines coloured by magnitude of velocity. ................................................................................... 64

Figure 59: Indo-Sail scale model tested at the Alster in Hamburg ................... 65

Figure 60: Achievable sail forces over apparent wind angle ............................ 66

Figure 61: Comparison of equivalent power, Dyna-Rig and Indo-Sail .............. 66

Figure 62: Kite application in a small cargo ship (www.skysails.com) .............. 67

Figure 63: Relative fuel saving for tanker of the case study ............................. 67

Figure 64: "Magnus” effect on a rotating cylinder. .......................................... 69

Figure 65: ENERCON E-Ship 1, 2010. ............................................................. 69

Figure 66: Potential application of Flettner rotor on a bulk carrier................... 70

Figure 67: Achievable thrust from three rotors over Apparent Wind Angle. ..... 70

Figure 68: Functional principle of a solar cell ................................................. 71

Figure 69: Potential Areas for the installation of PV-Units. .............................. 72

Figure 70: Power alternation for internal combustion engines (above) and for fuel cells (below). ........................................................................................... 74

Figure 71: Theoretical comparison between the Carnot Factor ad Fuel Cell Efficiency ....................................................................................................... 74

Figure 72: Construction of the "Alsterwasser". ................................................ 77

Figure 73: The FCS is placed on the main deck in a container. In service the system is stored inside the accommodation opening. ..................................... 77

Figure 74: Ragone overview ........................................................................... 80

Figure 75: Simplified ship propulsion system ................................................. 84

Figure 76: Simplified ship resistance decomposition....................................... 85

Figure 77: FO System Decomposition ............................................................. 86

Figure 78: HVAC System Decomposition ........................................................ 87